Food Biochemistry and Food Processing (Simpson/Food Biochemistry and Food Processing) || Equid Milk: Chemistry, Biochemistry and Processing

P1: SFK/UKS P2: SFK

BLBS102-c26 BLBS102-Simpson March 21, 2012 13:51 Trim: 276mm X 219mm Printer Name: Yet to Come

26Equid Milk: Chemistry, Biochemistry

and ProcessingT. Uniacke-Lowe and P.F. Fox

IntroductionEquid EvolutionEquid DomesticationRuminants and Non-RuminantsWhy Equid Milk in Human Nutrition?Production of Equid Milk

Composition of Equid MilkFactors that Affect the Composition

of Equid MilkStage of LactationEffect of Equid Breed on Milk Composition

ProteinsCaseins

αS1-CaseinαS2-Caseinβ-Caseinκ-Casein

Glycosylation of κ-CaseinHydrolysis of κ-Casein

Equid Casein MicellesStability of Equid Casein MicellesEnzymatic Coagulation of Equid MilkAcid-induced Coagulation of Equid MilkHeat-induced Coagulation of Equine MilkStability of Equine Milk to Ethanol

Whey Proteinsβ-Lactoglobulinα-LactalbuminImmunoglobulinsLactoferrinWhey Protein Denaturation

Digestibility of Equid MilkTotal Amino AcidsNon-protein nitrogen

Free Amino AcidsBioactive PeptidesHormones and Growth FactorsAmyloid A

Indigenous EnzymesLysozymeOther Indigenous Enzymes in Equine Milk

CarbohydratesLactose and GlucoseOligosaccharides

LipidsFatty Acids in Equid MilksStructure of TriglyceridesEquid Milk Fat Globules and MFGMRheology Equid Milk Fat

Stability of Equine Milk FatVitaminsMinerals

Macro-ElementsTrace Elements

Physical Properties of Equid MilkDensityRefractive IndexpHFreezing PointViscosityColour

Processing of Equid MilkKoumissOther Products from Equid Milk

Nutritional and Biomedical Properties of Equid MilkCow Milk Protein AllergyCross-Reactivity of Milk Proteins

SummaryReferences

Abstract: The characteristics of equid milk of interest in humannutrition include a high concentration of polyunsaturated fatty acids,a low cholesterol content, high lactose and low protein levels, aswell as high levels of vitamins A, B and C. Compared to bovinemilk, the protein fraction of equid milk contains proportionally less

Food Biochemistry and Food Processing, Second Edition. Edited by Benjamin K. Simpson, Leo M.L. Nollet, Fidel Toldra, Soottawat Benjakul, Gopinadhan Paliyath and Y.H. Hui.C© 2012 John Wiley & Sons, Inc. Published 2012 by John Wiley & Sons, Inc.

491

P1: SFK/UKS P2: SFK


492 Part 4: Milk

casein and more whey proteins. The low fat and unique fatty acidprofile of both equine and asinine milk result in low atherogenic andthrombogenic indices. The high lactose content of equid milk givesgood palatability and improves the intestinal absorption of calciumthat is important for bone mineralisation in children. The renal loadof equid milk, based on levels of protein and inorganic substances,is equal to that of human milk, a further indication of its suitabilityas an infant food. The concentration of lysozyme, lactoferrin andn-3 fatty acids is exceptionally high in equid milk, suggesting apotential anti-inflammatory effect. Equine and asinine milk can beused for their prebiotic and probiotic activity and as alternatives forinfants and children with cow milk protein allergy and other foodintolerances.

While the gross composition of equine and asinine milk has beenreasonably well established and the caseins have been fractionatedand well characterised, the presence of κ-casein remains a con-tentious issue and there is little or no information on the physico-chemical properties of equid milk. The only significant product fromequine milk is the fermented product, koumiss.

The composition of equid milk suggests a product with inter-esting nutritional characteristics with potential use in dietetics andtherapeutics, especially in diets for the elderly, convalescent andnewborns, but the lack of scientific research must be addressed todevelop the potential of equine and asinine milk in the health andnutritional markets.

INTRODUCTIONApproximately one-third of all mammalian genera are herbi-vores, more than half of which belong to two orders, (1) thePerissodactyla (odd-toed ungulates (hoofed animals)) and (2)the Artiodactyla (even-toed ungulates) (Savage and Long 1986).The horse and donkey belong to the order Perissodactyla thathas three families: (1) Equidae (nine species of horses, don-key and zebras), (2) Tapiridae (four species of tapirs) and (3)Rhinocerotidae (five species of rhinoceros). Uniquely amongspecies, all Equidae can interbreed, but the hybrid offspringare almost always infertile because horses have 64 chromo-somes and donkeys have 62 chromosomes (Trujillo et al. 1962).Zebras have between 32 and 46 chromosomes, depending onbreed (Burchelli’s zebra, Equus burchelli, has 44 chromosomes,Carbone et al. (2006) and viable hybrids with donkeys havebeen produced where gene combinations have allowed forembryonic development to birth. Differences in chromosomenumbers between horses and zebras are most likely due tohorses having two longer chromosomes that contain a singlegene content compared to four zebra chromosomes (Benirschkeet al. 1964).

This chapter presents a brief historical overview of some as-pects of the domestication of equid species, a review of the chem-istry and biochemistry of the principal constituents of equinemilk and, to a lesser extent, of asinine milk, with comparativedata for bovine and human milk. The technological propertiesof equid milk, with reference to processing of the milk are ex-amined and, finally, a synopsis of the nutritional and biologicalsignificance of equid milks in the human diet is presented.

Equid Evolution

Perissodactyla species evolved during the early Eocene (∼55million years ago) and were the dominant ungulate order un-til approximately 15 million years ago when numbers declined,probably due to climatic factors rather than competition withemerging Artiodactyla species (including the Ruminantia fam-ily) (MacDonald 2001). Nevertheless, wild equids ranged in vastherds across the grasslands of the Northern Hemisphere andin South America until the end of the ice age, approximately10,000 years ago. Wild horses were the chief prey of increasinghuman populations and numbers decreased until they becameextinct in North America approximately 7000 years ago and inEurope, herds of horses were pushed eastwards into Central Asiawhere the last few Przewalski horses (Equus ferus przewalski)survived until the twentieth century (Clutton-Brock 1992). Itis generally assumed that equids were domesticated approxi-mately 5000 years ago but archaeologists believe man did notride horses until approximately 1000 bc, although horses re-placed the ox as draft animals approximately 2000 bc (Clutton-Brock 1992). One of the early records of the domestic horsein Western Europe comes from horse remains found at the lateNeolithic site at Newgrange, County Meath, Ireland (Clutton-Brock 1992). Outram (2009) demonstrated domestication of thehorse in the Eneolithic Botai culture of Kazakhstan approxi-mately 3500 years ago and analysis of organic residues based oncharacterisation of stable isotopes of carbon, δ13C (which allowsdifferentiation of non-ruminant and ruminant carcass and dairyfats) and deuterium δD, values of fatty acid analysis revealedprocessing of mares’ milk and meat in ceramics at that time.The donkey (Equus asinus) is believed to have evolved from theNubian and Somalian sub-species of African wild asses approx-imately 4000–5000 years ago and has since been an integral partof human life as a pack and riding animal. Today, perissodactylsare a poor second to artiodactyls in terms of numbers of species,geographical distribution, variety of form and ecological diver-sity (MacDonald 2001). For further details on equid evolutionand genetic lineages of domestic horse species, see Vila et al.(2001) and references therein.

Equid Domestication

Only five major species of large, plant-eating mammals havebeen widely domesticated: sheep, goat, cattle, pig and horse.Nine additional minor species have been domesticated but arerestricted to certain geographical areas: Arabian and Bactriancamels, llama, alpaca, donkey, reindeer, water buffalo, yak,Bali cattle of Southeast Asia and the gaur (mithan) of Indiaand Burma (Bruns 1999). Despite taxonomic congruity and be-havioural similarities between equid species, only two equidshave been domesticated: the horse (Equus ferus) and the donkey(Equus asinus africanus) (Clutton-Brock 1992).

Because milk and products derived from it provide up to 30%of dietary protein in developed countries, to meet this demand,man has genetically improved some species for milk production(Mercier 1986). Domesticated animals have been modified fromtheir wild ancestors through being kept and selectively bred for

P1: SFK/UKS P2: SFK


26 Equid Milk: Chemistry, Biochemistry and Processing 493

use by humans who control the animals’ breeding and feeding.Domestication of dairy species has meant that different breedsof farm animals, for example dairy cows and goats, can be devel-oped and maintained to optimise certain hereditary factors, forexample, length of lactation, period of gestation and milk yield,as well as protein and fat levels in the milk produced. Geneticselection of breeds of horse and donkey for milk production hasnot occurred, as yet, and consequently there is high variabilityfor milk yield and length of lactation, as well as high individualvariability.

In some regions of the world, for example Mongolia andSouthern Russia, Hungary, France and Belgium, the horse and,less frequently, the donkey, is an important source of meatand milk. Steppe Mongols, forest-steppe Kazakhs, the Hadzahunter–gatherers of Tanzania and the urban French all regardhorses as a valuable food source and believe that horse flesh andmilk have special nutritional and medicinal attributes (Levine1998). The use of donkeys as a dairy species can be tracedback to Roman times when the nutritional value of its milk andbeneficial properties in skin care were first recognised (Salimei2011).

The importance of the horse in leisure activities, especiallyracing, has led to the scientific breeding and nutrition of horses,including foals, and has created the need to characterise thecomposition and properties of equine milk, which are now rela-tively well known. The fact that the horse is spread throughoutthe world, has been domesticated, is relatively easily handled,and produces large amounts of milk, which can be obtained rel-atively easily, makes equine milk a relatively easy subject forresearch. Although the donkey is now less widely distributedthan the horse, its milk can be obtained readily. The zebra hasnot been domesticated and although it is widespread in the wildin Africa and in captivity elsewhere, apparently it is very dif-ficult to obtain zebra milk, even from captive animals. Studiesof the social behaviour of equids have shown that there are nointrinsic reasons why the zebra has never been domesticated butit is believed that the peoples of Africa may have had culturalrather than biological reasons not to use zebras as pack animals(Clutton-Brock 1992).

Ruminants and Non-Ruminants

The horse and donkey are non-ruminant herbivores and digestportions of their feed first enzymatically in a foregut and thenferment it in a very large sacculated hind-gut. Limited digestionoccurs in the equid stomach which liquefies incoming feed andsecretes gastric acid and pepsin to initiate breakdown of feedcomponents. The equid digestive system is designed to processsmall amounts of food frequently (Sneddon and Argenzio 1998).Equids rarely fast for more than 2–4 hours at a time and natu-rally forage for 16–18 hours/day. Horses in the wild roam widely,grazing both day and at night on immature, easily digested foodand exhibit few digestive problems in comparison to domesti-cated horses (Sellnow 2006). Unlike ruminants, and in keepingwith their status as prey animals, horses do not require periodsof rest to stop and ruminate. Ruminants have very efficient di-gestive systems with microbial breakdown of fibrous food at the

start of the gastrointestinal tract and nutrient absorption alongthe entire intestine. Ruminants can digest fiber and carbohydratemore completely than any other species.

Donkeys, like horses, generally survive on a diet high in fiberand low in soluble carbohydrate and protein but while donkeysare not as efficient as ruminants at digesting cell wall compo-nents, they are far more efficient than horses (Izraely et al. 1989).Donkeys are capable of consuming large amounts of forage andgain more digestible energy from it than even goats fed a simi-lar diet (Smith and Sherman 2009). The donkey achieves this bysubstantially increasing its forage intake rate to compensate for alow-quality diet. A donkey uses its narrower muzzle and prehen-sile lips for greater selectivity of its food, thereby, maximisingfeed quality rather than quantity (Aganga et al. 2000). Don-keys have a much lower water requirement per unit weight thanany other mammal, except the camel, and can rehydrate quicklywith large volumes of water without complications. Donkeyscan work while suffering from severe dehydration by reducingwater and energy turnover rates, while maintaining feed intakeand its plasma volume can be maintained by drawing on thesubstantial fluid reservoir in the hind-gut (Sneddon et al. 2006).The zebra, on the other hand, needs a constant source of waterfor survival (Aganga et al. 2000).

Why Equid Milk in Human Nutrition?

The benefits of equine milk for human health is an ancient ideaand there is much literature from the former Soviet Union onthis subject, although it is now accepted that the results of ex-perimental work are dubious (Doreau and Martin-Rosset 2002).Because equine milk resembles human milk in many respectsand is claimed to have special therapeutic properties, it is be-coming increasingly important in Western Europe, especially inFrance, Italy, Hungary and the Netherlands. Equine milk (andkoumiss, fermented equine milk) is often used for the treatmentof a myriad of ailments including anaemia, nephritis, diarrhoea,gastritis disorders, cardiovascular disease and in post-operativecare, as well as for stimulation of the immune system (Lozovich1995). In Mongolia, where koumiss is the national drink, peoplehave a saying that ‘kumys cures 40 diseases’ (Levine 1998). InItaly, equine milk has been recommended as a possible substi-tute for bovine milk for allergic children (Curadi et al. 2001).Equine milk is considered to be highly digestible, rich in es-sential nutrients and possesses an optimum whey protein:caseinratio, making it very suitable as a substitute for bovine milk inpaediatric dietetics. Estimates suggest that more than 30 millionpeople drink equine milk regularly, with this figure increasingsignificantly annually (Doreau and Martin-Rosset 2002).

The use of asinine milk by humans for alimentary andcosmetic purposes has been popular since Egyptian antiquity.Cleopatra is reputed to have bathed daily in asinine milk andkept a herd of 700 to fill her bath. Hippocrates (460–370 bc) wasa strong advocate of the use of asinine milk as a medicine andused it to cure many ailments including, liver disease, oedema,nosebleed, poisoning and wounds. Today, asinine milk is con-sumed mainly in countries where donkeys were traditionallybred, Asia, Africa and Eastern Europe but more recently it has

P1: SFK/UKS P2: SFK


494 Part 4: Milk

been used successfully as a substitute for human milk in West-ern Europe (Vincenzetti et al. 2008) and is the milk of choice inItaly for children with severe immunoglobulin E (IgE)-mediatedcows’ milk allergy. (Businco et al. 2000).

Production of Equid Milk

The production of equine milk and the factors that affect it havebeen the subject of several reviews, including Doreau and Boulot(1989), Doreau et al. (1990), Doreau (1994), Doreau and Martin-Rosset (2002) and Park et al. (2006) and, therefore, is consideredonly briefly here.

World milk production was approximately 695 million tonnesin 2009, of which 84% was bovine, 13% buffalo and 3% sheep,goat and other species (IDF 2009). Statistics for milk productionfrom species other than cows and buffalo are not very reliableand are available only from countries where these milks areprocessed industrially although it is accepted that, while bovinemilk production figures have changed little in recent years, theproduction of milk from buffalo, camels, horses and donkeys isincreasing. In Europe, it is estimated that about 1–1.3 millionlitres of equine milk are currently produced per annum but incountries such as Mongolia the figure is considerably higher,probably approximately 9 million litres.

A decade ago, equine milk was produced only in isolatedsmall holdings in parts of Eastern Europe and Mongolia, butnow there are large-scale operations in France, Belgium, Ger-many, Austria and the Netherlands. Asinine milk is produced inlarge donkey farms in Italy, France, Spain, Belgium, Xinjiangand Shanxi provinces of China, Ethiopia and Pakistan (Salimei2011).

For equine milk production, milking begins when the foalreaches approximately 8 weeks and is eating some hay and grass.The mare is separated from the foal by day and milked approx-imately five times at intervals of about 2.5 hours and produces1–1.5 L of milk at each milking. At night, the foal feeds freely(van Laar, Orchid’s Paardenmelkerij, the Netherlands, personalcommunication). Thus, milk production is very labour intensiveand expensive, with the result that equine milk is priced as a del-icacy, typically, €11/L. Milking schedules are similar for asininemilk, but the yield is lower than that of the horse, approximately350–850 mL of milk per milking, depending on several fac-tors including: foal and mare management, milking procedure,stage of lactation, body size and condition and feeding regime(Salimei 2011). Mastitis is rarely a problem with equids andoccurs only if teat injury occurs during milking. Furthermore,equid species appear to be relatively resistant to brucellosis andtuberculosis, which is advantageous from a dairy farming pointof view (Stoyanova et al. 1988).

COMPOSITION OF EQUID MILKWith the exception of the major domesticated dairy species andhumans, information on milk composition is poor and of >4500mammalian species in existence, milk compositional data areavailable for approximately 200 species, of which, data for onlyapproximately 55 species appears to be reliable. The milk of all

mammals contains the same principal components: water, salts,vitamins, fats, carbohydrate and proteins, but these constituentsdiffer significantly both quantitatively and qualitatively betweenspecies (Table 26.1), although species from the same taxonomicorder, for example equids, produce milk of similar composition(Table 26.1). Equid milk is similar in composition to humanmilk but considerably different from that of other dairy mam-mals, for example cow, buffalo, sheep, goat, camel, llama andyak (Table 26.1). Why equid milk is so similar in macro compo-sition to that of human milk is unclear, especially as equids andhumans are phylogenetically distantly related. Inter-species dif-ferences in milk composition reflect very divergent patterns ofnutrient transfer to the young and presumably reflect adaptationsin maternal rearing of offspring to physiological constraints andenvironmental conditions (Oftedal and Iverson 1995). In all sit-uations, lactation must be effective in providing nourishment tothe offspring without over-taxing maternal resources.

The protein content of milk varies considerably betweenspecies and reflects the growth rate of the young. Bernhart(1961) found a linear correlation between the percent of caloriesderived from protein and the logarithm of the days to doublebirth weight for 12 mammalian species. For humans, one of theslowest growing and slowest maturing species, it takes 120–180days to double birth weigh and only 7% of calories come fromprotein. In contrast, carnivores can double their birth weigh inas little as 7 days and acquire >30% of their energy from pro-tein. Equid species take between 30 and 60 days to double theirbirth weight and, like humans, have an exceptionally low levelof protein in their milk (Table 26.1). The high metabolic needsof the foal are met through frequent feeding. Equid milks havea significantly lower energy value than human milk, owing totheir low fat content (Table 26.1).

The fat content of milk across species shows large variationand ranges from approximately 0.6% for some breeds of donkeyand less than 1% for rhinoceros to approximately 50% in the milkof some seals. High-fat milks are important for some species,for example desert mammals, when maternal water economy isimportant and energy needs to be transferred to the young in anefficient manner. Equid milk has one of the highest contents ofcarbohydrate, which is similar to that of human milk. In equidspecies, lactation lasts naturally for approximately 7 months.

Factors that Affect the Compositionof Equid Milk

Stage of Lactation

Within species, the stage of lactation is the most importantdeterminant of milk composition and the difference betweencolostrum and mid-lactation milk shows the most significantdifference but absolute values and the direction of changevary among species (Casey 1989). Shorty after parturition, themammary gland produces colostrum that is richer in dry mat-ter, proteins, fat, vitamins and minerals (except calcium andphosphorus) but poorer in lactose than mature milk. One of themajor biological benefits of colostrum is the presence of Igs,IgA, IgM and IgG, and high levels of some enzymes, including

P1: SFK/UKS P2: SFK



Table 26.1. Gross Composition of the Milk of Equid Species and Some Dairy Species, with Human and OtherSelected Species Included for Comparison

Gross Days toCasein: Energy Double

Total Whey (kJ.kg−1 BirthSpecies Solids Protein Ratio Fat Lactose Ash or kJ.L−1) Rate

aHorse (Equus caballus) 102.0 21.4 1.1:1 12.1 63.7 4.2 1883 40–60aDonkey (Equus africanus

asinus)88.4 17.2 1.28:1 14.0 68.8 3.9 1966 30–50

aMountain zebra (Equuszebra hartmannae)

100.0 15.6 – 10.2 69.0 3.0 1800 –

aPlains zebra (Equusburchelli)

113.0 16.3 – 22.0 70.0 4.0 2273 –

aPrzewalski horse (Equuscaballus przewalski)

105.0 15.5 1.1:1 15.0 67.0 3.0 1946 –

bCow (Bos taurus) 127.0 34.0 4.7:1 37.0 48.0 7.0 2763 30–47aBuffalo (Bubalus bubalis) 172.0 46.5 4.6:1 81.4 48.5 8.0 4644 48–50bSheep (Ovis aries) 181.0 55.9 3.1:1 68.2 48.8 10.0 4309 10–15aGoat (Capra hircus)h 122.0 35.0 3.5:1 38.0 41.0 8.0 2719 12–19bCamel (Camelus

dromedarius)124.7 33.5 1.68:1 38.2 44.6 7.9 2745 250

b Llama (Llama glama) 131.0 34.0 3.1:1 27.0 65.0 5.0 2673 120aYak (Bos grunniens) 160.0 42.3 4.5:1 56.0 52.9 9.1 3702 60aMan(Homo sapiens) 124.0 9.0 0.4:1 38.0 70.0 2.0 2763 120–180bPig (Sus scrofa) 188.0 36.5 1.4:1 65.8 49.6 10.0 3917 9bRabbit (Oryctolagus

cuniculus)328.0 139.0 2.0:1 183.0 21.0 18.0 9581 4–6

bBlue whale (Balaenopteramusculus)

550.0 119.0 2.0:1 409.0 13.0 14.0 17614 10

bNorthern Fur Seal(Callorhinus ursinus)

633.0 103.0 1.1:1 507.0 1.0 5.0 20836 5

aRat (Rattus norvegicus) 210.0 84.0 3.2:1 103.0 26.0 13.0 5732 2–5aElephant (Loxodonta

africana africana)176.9 47.3 0.61:1 60.7 38.8 7.0 3975 100–260

aRhinoceros (Ceratotheriumsimum)

77.5 16.2 0.22:1 7.4 61.0 3.0 1589 25–35

Source: Modified from Uniacke et al. 2010.Values are expressed as ag.kg−1 or bg.L−1 Milk.

catalase, lipase and proteinase. Figure 26.1 shows the affect oflactation on the main constituents of equine milk and indicatesa very rapid transition from equine colostrum to mature equinemilk, that is within the first 24 hours of lactation. The concen-tration of protein in equine milk is very high, >15 g/100 g milk,immediately post-partum, but decreases rapidly to <4 g/100 gmilk after 24 hours of lactation and to less than 2 g/100 g milkafter 4 weeks of lactation (Fig. 26.1A). The casein to whey pro-tein ratio in equine colostrum is 0.2:1 immediately post-partumand this changes to approximately 1.1:1 within 1 week. The pro-tein content of bovine milk decreases during the first 3 monthsof lactation, but increases subsequently (Walstra et al. 2006a).The concentration of lactose in equine more than doubles dur-ing the first 24 hours (Fig. 26.1C), and this is also observed forbovine milk (Walstra et al. 2006a). The concentration of lactosein equine milk subsequently increases steadily throughout fur-

ther lactation (Fig. 26.1C), a trend that is different from that forbovine milk in which lactose content decreases progressively(Walstra et al. 2006a). Close agreement is observed betweenthe data of various studies for the level of protein (Fig. 26.1A)and lactose (Fig. 26.1C) in equine milk but considerable differ-ences are reported for the lipid content of equine milk betweendifferent studies (Csapo et al. 1995, Doreau et al. 1986). Thismay be due to an increase in the fat content of equine milk thatoccurs during a milking session and, in some cases, the use ofthe hormone, oxytocin, which promotes complete evacuationof the udder (Doreau et al. 1986). Hence, the volume of milkdrawn and the degree of evacuation of the udder will signifi-cantly influence the lipid content of the milk and thus explaindifferences in lipid content observed between different studies.However, all studies indicated in Figure 26.1B show the sametrend, that is a decrease in the lipid content of equine milk with

P1: SFK/UKS P2: SFK


496 Part 4: Milk

Pro

tein

(g

/100

g m

ilk)

0

2

416

18

20L

ipid

s (g

/100

g m

ilk)

0

1

2

3

4

Stage of lactation (d)

1801501209060300

Lac

tose

(g

/100

g m

ilk)

0

2

4

6

8

(A)

(B)

(C)

Figure 26.1. Influence of lactation stage on the concentration of (A)protein, (B) lipids, or (C) lactose in equine milk. (Data from Ullreyet al. (1966, •), Mariani et al. (2001, ◦) Smolders et al. (1990, �)and Zicker and Lonnerdal (1994, ∇).)

advancing lactation (Fig. 26.1B), whereas that the lipid contentin bovine milk shows a distinct minimum after approximately3 months of lactation (Walstra et al. 2006a). The fat content ofasinine milk increases from approximately 0.5% to 1.5% fromdays 15 to 105 but decreases sharply thereafter (Guo et al. 2007,Salimei et al. 2004). Piccione et al. (2008) reported a decrease

in fat in the milk of Ragusana donkeys throughout lactation andobserved a daily rhythmicity, similar to that found in bovine andhuman milk, for the levels of fat, lactose and protein in asininemilk, with fat and lactose peaking at night and protein reachinga maximum level during the daytime.

Effect of Equid Breed on Milk Composition

Data in the literature are not conclusive as to whether or notthe breed of mare has an effect on the concentration of proteinin milk. Kulisa (1977), Doreau et al. (1990), Csapo-Kiss et al.(1995) and Csapo et al. (1995) reported no effect of breed on theconcentration of proteins or lipids in equine milk throughout lac-tation. On the other hand, Boulot (1987) and Formaggioni et al.(2003) have reported significant differences in protein contentbetween breeds. Civardi et al. (2002), who compared Arabian,Haflinger, Trotter and Norico breeds, found that Norico milkhad significantly lower α-lactalbumin (α-La), highest lysozyme(Lyz) and β-lactoglobulin (β-Lg) and highest thermal resistanceof the breeds studied. Pelizzola et al. (2006), who compared themilk of Haflinger, Quarter horse, Sella/Salto and Rapid HeavyDraft, found that Quarter horse milk had significantly higherconcentrations of the main constituents and higher concentra-tions of linoleic and α-linolenic fatty acids (ALA) than in themilk of the other species.

Asinine milk shows variability in fat content among breedsand is reported to be as low as 0.4% for Martina Franca mares,0.6% in Ragusana mares and as high as 1.7% in Jiangyue don-keys (for these donkeys an increase from 0.5% to 1.7% wasrecorded in the fat content of the milk over 180 days of lac-tation) (Guo et al. 2007). Milk yield is significantly lower forJiangyue donkeys than for Martina Franca and Ragusana breedsand the protein pattern of Jangyue milk is significantly differentfrom the other breeds (Guo et al. 2007)

PROTEINSWhile the protein content of mature equid milk is lower thanthat of bovine milk, there is a strong qualitative resemblance, theprincipal classes of proteins, that is caseins and whey proteinsare similar in both milks. However, while the caseins are thepredominant class of proteins in bovine milk (∼80% of totalmilk protein), equid milk contains less casein and more wheyproteins. The distribution of casein and whey proteins in equidmilk is shown in Table 26.2, with comparative data for bovineand human milk.

Caseins

About 80% of the proteins in bovine milk are caseins that areprimarily a source of amino acids, calcium, phosphate and bioac-tive peptides for neonates (Shekar et al. 2006). The low-caseinconcentration in mature equine milk (∼55% of total protein) hasmany implications that will be discussed later. The traditionalmethod for separating caseins from whey proteins is isoelectricprecipitation of the caseins at pH approximately 4.6. The ca-sein fraction of most milks consists of four gene products: αs1-,

P1: SFK/UKS P2: SFK



Table 26.2. Concentration of Caseins and Whey Proteins (g. kg−1) in Equine, Asinine,Human and Bovine Milk

Equine Asinine Human Bovine

Total casein 13.56 7.8 2.4 26αs1-casein 2.4 Identified 0.77 10.7αs2-casein 0.20 Unknown – 2.8β-casein 10.66 Identified 3.87 8.6κ-casein 0.24 Unknown 0.14 3.1γ -casein Identified Unknown – 0.8Total whey protein 8.3 5.8 6.2 6.3β-lactoglobulin 2.55 3.3 – 3.2α-lactalbumin 2.37 1.9 2.5 1.2Serum albumin 0.37 0.4 0.48 0.4Proteose peptone – – – 0.8Immunoglobulins 1.63 1.30 0.96 0.80IgG1,2 0.38 0.03 0.65IgA 0.47 0.96 0.14IgM 0.03 0.02 0.05Lactoferrin 0.58 0.37 1.65 0.10Lysozyme 0.87 1.00 0.34 126 × 10−6

NPN (mg.L−1) 375 455 454 266Casein micelle size (nm) 255 ∼100–200 64 182

Source: Modified from Uniacke et al. 2010, with asinine data from Guo et al. 2007 and Salimei et al. 2004.NPN, non-protein nitrogen.

αs2-, β- and κ-caseins, of which the first three are calcium sen-sitive. All caseins lack secondary structure, which led Holt andSawyer (1993) to consider them as rheomorphic proteins. Thelack of secondary structure may be attributed, at least partially,to the relatively high level of proline residues in casein. As aresult, caseins do not denature or associate on heating (Paulsonand Dejmek 1990). The biological function of the caseins liesin their ability to form macromolecular structures, casein mi-celles, which transfer large amounts of calcium to the neonatewith a minimal risk of pathological calcification of the mam-

mary gland. The individual caseins will be discussed separatelyin the following sections with focus on their interactions to formcasein micelles and the colloidal stability thereof.

Fractionation and characterisation of individual equine ca-seins has been poorly researched to date in comparison to thoseof bovine milk and it had been reported that equine, and presum-ably asinine, caseins exhibit greater heterogeneity and a higherlevel of post-translational modifications than those of bovinemilk (Miranda et al. 2004). Table 26.3 shows the biochemicalproperties of individual casein proteins that are discussed later.

Table 26.3. Properties of Equine, Bovine and Human αs1-, β- and κ-Caseins

Protein Species

PrimaryAccessionNumbera

AminoAcid

ResiduesMolecular

Weight (Da) pH GRAVY bCysteineResidues

αs1-casein Equine Q8SPR1 205 24,614.4 5.47 −1.127 0Bovine P02662 199 22,974.8 4.99 −0.704 0Human P47710 170 20,089.4 5.17 −1.013 3

β-casein Equine Q9GKK3 226 25,511.4 5.78 −0.415 0Bovine P02666 209 23,583.2 5.13 −0.355 0Human P05814 211 23,857.8 5.33 −0.289 0

κ-casein Equine P82187 165 18,844.7 8.03 −0.313 2Bovine P02668 169 18,974.4 5.93 −0.557 2Human P07498 162 18,162.6 8.68 −0.528 1

Source: Modified from Uniacke et al. 2010.aPrimary accession number for the protein in SWISS-PROT database.bGrand average hydropathy (GRAVY) score using the scale of Kyte and Doolittle (1982).

P1: SFK/UKS P2: SFK


498 Part 4: Milk

αS1-Casein

The amino acid sequence of equine αs1-casein has been deducedfrom its cDNA sequence (Lenasi et al. 2003). The protein con-tains 205 amino acids and has a molecular mass of 26,614.4Da prior to post-translational modification, that is, it is con-siderably larger than its bovine or human counterpart (Table26.3). Two smaller isoforms of αs1-casein have been identifiedin equine milk, which probably result from the skipping of ex-ons during transcription (Miranda et al. 2004). Equine αs1-caseincontains six potential phosphorylation sites (Lenasi et al. 2003),five of which are in very close proximity (Ser75, Ser77, Ser79,Ser80, Ser81) and can thus form a phosphorylation centre, whichis important in the structure of casein micelles. Mateos et al.(2009a) determined the different phosphorylation levels of thenative isoforms of equine αs1-casein and identified 36 differ-ent variants with several phosphate groups ranging from twoto six or eight which, like equine β-casein, present a complexpattern on one dimensional and two dimensional electrophore-sis. Bovine αs1-casein contains eight or nine phosphorylationsites (Swaisgood 2003), which form two phosphorylation cen-tres (De Kruif and Holt 2003). Bovine αs1-casein contains threedistinct hydrophobic regions, roughly including residues 1–44,90–113 and 132–199 (Swaisgood 2003). These regions are char-acterised by positive values for hydropathy. Likewise, equineαs1-casein has three domains with a high hydropathy value, thatis, around residues 25–30, 95–105 and 150–205 and thereforeit probably has association properties similar to those of bovineαs1-casein. Furthermore, equine αs1-casein contains two regionswith very low hydropathy, that is, around residues 45–55 and125–135, which are expected to behave hydrophilically. Humanαs1-casein does not appear to have distinct hydrophobic regions.Overall, equine and human αs1-casein have comparable grandaverage hydropathy (GRAVY) scores, which are lower than thatof bovine αs1-casein (Table 26.3), indicating an overall higherhydrophobicity for the latter. GRAVY scores reflect the relativeratio of hydrophobic and hydrophilic amino acid residues in aprotein, with a positive value reflecting an overall hydrophobicand a negative value an overall hydrophilic nature of the protein.

Prior to the mid-1990s, it was generally assumed that hu-man milk contains mainly β- and κ-caseins with little or noαs-casein (Kunz and Lonnerdal 1990). A minor casein compo-nent has since been identified and is considered to be the humanequivalent of αs1-casein, although this identification highlightsseveral inconsistencies in comparison with the equivalent caseinin other species. Uniquely, human αs1-casein appears to con-tain at least two cysteine residues and exists as a multimer incomplex with κ-casein (Cavaletto et al. 1994, Rasmussen et al.1995). Johnsen et al. (1995) identified three cysteine residuesin human αs1-casein and provided a molecular explanation forαs1-κ-casein complex formation. Martin et al. (1996) provideddefinitive evidence for the presence of a functional αs1-caseinlocus in the human genome which is expressed in the mam-mary gland during lactation, while Sørensen et al. (2003) de-termined the phosphorylation pattern of human αs1-casein. Inbovine milk, αs1- casein is a major structural component of thecasein micelle and plays a functional role in curd formation

(Walstra and Jenness 1984). The relatively low level of αs1-casein in equine milk (Table 26.2), and similarly in human milk,may be significant and, coupled with the low protein content,could be responsible for the soft curd produced in the stom-ach of the infant or foal (Dr. Ursula Fogarty, National EquineCentre, Ireland – personal communication). Goat milk lackingαs1-casein has poor coagulation properties compared to milkcontaining αs1-casein (Clark and Sherbon 2000). Bevilacquaet al. (2001), who assessed the capacity of goat’s milk with a lowor high αs1-casein content to induce milk protein sensitisationin guinea pigs, found significantly less sensitisation in milk withlow αs1-casein. This may represent another important attributeof the low αs1-casein content of equine milk for use in humanallergology.

An αs1-like protein of approximately 31–33 kDa has beenidentified in asinine milk although Criscione et al. (2009) re-ported its absence in one Ragusana donkey under investigation.

αS2-Casein

The complete amino acid sequence of equine αs2-casein is un-known, but Ochirkhuyag et al. (2000) published the sequence ofthe N-terminal 15 amino acid residues (Lys-His-Lys-Met-Glu-His-Phe-Ala-Pro-Xaa-Tyr-Xaa-Gln-Val-Leu, where Xaa is anunknown amino acid). Only five of these amino acids were con-firmed by Miranda et al. (2004). Isoelectric focusing showed twomajor bands for equine αs2-casein, with isoelectric points in thepH range 4.3–5.1 (Ochirkhuyag et al. 2000). Bovine αs2-caseinis the most highly phosphorylated casein, usually containing 11phosphorylated serine residues, with lesser amounts containing10, 12 or 13 phosphate groups (Swaisgood 2003). There areno reports on the presence of αs2-casein in human milk. Usingthree different methods for protein identification, Criscione et al.(2009) could not detect αs2- or κ-casein in asinine milk.

β-Casein

The amino acid sequence of equine β-casein, derived fromthe cDNA, has been reported by Lenasi et al. (2003), and re-vised by Girardet et al. (2006) with the insertion of eight aminoacids (glutamic acid (Glu27) to Lys34). The theoretical molecularmass of this 226 amino acid polypeptide is 25,511.4 Da (Table26.3). Bovine and human β-casein contain 209 and 211 aminoacid residues, respectively (Table 26.3). Two smaller variantsof equine β-casein, which probably result from casual exon-skipping during transcription, were reported by Miranda et al.(2004). The 28 C-terminal amino acids contain seven poten-tial phosphorylation sites (Ser9, Ser15, Ser18, Ser23, Ser24, Ser25,Ser28) and multiple-phosphorylated isoforms of equine β-caseincontaining three to seven phosphoserine residues have been re-ported, with the isoelectric point varying from pH 4.74–5.30(Girardet et al. 2006, Mateos et al. 2009b). Bovine β-casein,which contains four or five phosphorylated serine residues,has an isoelectric point of 5.0–5.5 (Swaisgood 2003). Humanβ-casein has up to six levels of phosphorylation, that is, 0, 1, 2, 3,4 or 5 phosphorylated serine residues (Sood and Slattery 2000).

P1: SFK/UKS P2: SFK



Equine, bovine and human β-casein have a very hydrophilicN-terminus, followed by a relatively random hydropathy distri-bution in the rest of the protein, leading to an amphiphilic proteinwith a hydrophilic N-terminus and a hydrophobic C-terminus.In equine sodium caseinate, the Lys47–Ile48, bond of β-caseinis hydrolysed readily by bovine plasmin, whereas no cleavageof the corresponding bond, Lys48–Ile49 in bovine β-casein hasbeen shown (Egito et al. 2003). In bovine β -casein, Lys28-Lys29

is readily cleaved by plasmin but the equivalent, Lys28–Leu29,in equine β-casein is insensitive (Egito et al. 2002). Otherplasmin cleavage sites in equine β-casein are Lys103–Arg104,Arg104–Lys105 and Lys105–Val196 (Egito et al. 2002). Equineβ-casein is readily hydrolysed by chymosin at Leu190–Tyr191

(Egito et al. 2001).Equine β-casein and equine α-La undergo spontaneous

deamidation under physiological conditions at Asn135–Gly136

and Asn45–Gly46, respectively (Girardet et al. 2004), which hasbeen reported also for canine milk Lyz (Nonaka et al. 2008) andhuman lactoferrin (Lf) (Belizy et al. 2001) but not, to our knowl-edge, for bovine or human β-casein or α-la. Recent research hasshown that temperature may be an important factor control-ling the spontaneous deamidation process and at 10◦C, the phe-nomenon is strongly reduced (Mateos et al. 2009b). Spontaneousdeamidation represents an important modification of equine milkproteins under certain conditions where bovine milk proteins,which do not contain a potential site for deamidation, remainunaffected. Equine Lf also contains the Asn–Gly sequence andmay be susceptible to spontaneous deamidation (Girardet et al.2006).

Unique to equine milk and apparently absent from the milkof other species, including ruminants, is a low-molecular weight(MW) multi-phosphorylated β-casein variant which accountsfor 4% of the total casein (Miclo et al. 2007). This short pro-tein (94 amino acid residues) is the result of a large deletion(residues 50–181) from full-length equine β-casein. No spon-taneous deamidation of this low-MW form of β-casein hasbeen found. Multi-phosphorylated isoforms of β-casein, approx-imately 34–35.4 kDa, have been identified in asinine milk butno further characterisation has been reported to date (Criscioneet al. 2009).

κ-Casein

The presence of κ-casein in equine milk was an issue of de-bate for several years, with several authors (Visser et al. 1982,Ono et al. 1989, Ochirkhuyag et al. 2000) reporting its absence.However, other studies (Kotts and Jenness 1976, Malacarne et al.2000, Iametti et al. 2001, Egito et al. 2001) showed its presence,albeit at a low concentration. The primary structure of equineκ-casein has been derived (Iametti et al. 2001, Lenasi et al. 2003,Miranda et al. 2004); it contains 165 amino acids residues, thatis four less than bovine κ-casein but three more than humanκ-casein (Table 26.3). The MW of equine κ-casein, prior topost-translational modification, is 18,844.7 Da. Equine and hu-man κ-casein have a considerably higher isoelectric pH thanbovine κ-casein (Table 26.3), and they have a net positive chargeat physiological pH, whereas bovine κ-casein has a net negative

charge. The GRAVY score of bovine κ-casein is considerablylower than that of equine κ-casein (Table 26.3), indicating thatthe latter is more hydrophilic. Bovine κ-casein is characterisedby a hydrophilic C-terminus, which is very important for themanner in which bovine casein micelles are stabilised, but acomparison of the hydropathy distribution of bovine and equineκ-caseins indicates that the C-terminus of equine κ-casein isfar less hydrophilic, particularly as a result of the absence of astrong hydrophilic region at residues 110–120. Human κ-caseinappears to be more like equine than bovine κ-casein in termsof the distribution of hydropathy along the polypeptide chain.Studies on asinine milk have not found κ-casein (Vincenzettiet al. 2008, Chianese et al. 2010).

Glycosylation of κ-Casein κ-Casein, the only glycosylatedmember of the casein family, exhibits microheterogeneity dueto the level of glycosylation (Saito and Itoh 1992). Tri- or tetra-saccharides consisting of N-acetylneuraminic acid (NANA),galactose and N-acetylgalactosamine are attached to κ-caseinvia O-glycosidic linkages to threonine residues in the C-terminalportion of the molecule (the glycomacropeptide region). Abouttwo-thirds of bovine κ-casein molecules are glycosylated at oneof six threonyl residues, that is Thr121, Thr131, Thr133, Thr135,Thr136 (only in bovine κ-casein variant A) or Thr142 (Pisano et al.1994); Ser 141 is also a potential glycosylation site (Kanamoriet al. 1981). Human κ-casein has seven glycosylation sites,Thr113, Thr123, Thr128, Thr131, Thr137, Thr147 and Thr149 (Fiatet al. 1980). Although no direct information is available, lectin-binding studies indicate that equine κ-casein is glycosylated(Iametti et al. 2001), possibly at residues Thr123, Thr127, Thr131,Thr149 and Thr153 (Lenasi et al. 2003) (these glycosylation sitesare not fully in agreement with those proposed by Egito et al.(2001)). To date, no non-glycosylated κ-casein has been identi-fied in equine milk (Martuzzi and Doreau 2006).

κ-Casein is located mainly on the surface of the casein mi-celles and is responsible for their stability (Walstra 1990). Thepresence of a glycan moiety in the C-terminal region of κ-caseinenhances its ability to stabilise the micelle, by electrostatic repul-sion, and may increase the resistance by the protein to proteolyticenzymes and high temperatures (Minkiewicz et al. 1993, Dzi-uba and Minkiewicz 1996). Biologically, NANA residues haveantibacterial properties and act as a bifidogenic factor (Dziubaand Minkiewicz 1996). κ-Casein is thought to play a major rolein preventing the adhesion of Helicobacter pylori to human gas-tric mucosa (Stromqvist et al. 1995). It is likely that heavily gly-cosylated κ-casein provides some protection to breast-feedinginfants due to its carbohydrate content which may be importantespecially as H. pylori infection is occurring at an increasinglyyounger age (Lonnerdal 2003).

Hydrolysis of κ-Casein The hydrolysis of bovine κ-caseinby chymosin at Phe105–Met106 leads to the production ofthe hydrophobic N-terminal para-κ-casein and the hydrophilicC-terminal caseinomacropeptide (CMP) (Walstra and Jenness1984). Chymosin hydrolyses the Phe97–Ile98 bond of equine κ-casein (Egito et al. 2001) and slowly hydrolyses the Phe105–Ile106

bond of human κ-casein (Plowman et al. 1999). However, as

P1: SFK/UKS P2: SFK


500 Part 4: Milk

Table 26.4. Properties of Equine, Bovine and Human Para-κ-Casein and Caseinomacropeptide

Protein Species Residues

AminoAcid

ResiduesMolecularWeight (Da) pI GRAVYa

Para-κ-casein Equine 1–97 97 11,693.3 8.96 −0.675Bovine 1–105 105 12,285.0 9.33 −0.617Human 1–105 97 11,456.9 9.63 −1.004

CMP Equine 98–165 68 7,169.3 4.72 0.203Bovine 106–169 63 6,707.4 4.04 −0.370Human 106–162 65 6,723.7 4.24 0.182

aGrand average hydropathy (GRAVY) score using the scale of Kyte and Doolittle (1982).Source: Modified from Uniacke et al. 2010.CMP, C-terminal caseinomacropeptide.

summarised in Table 26.4, the CMPs released from equine andhuman κ-caseins are considerably less hydrophilic than bovineCMP. The sequence 97–116 of κ-casein is highly conservedacross species, suggesting that the limited proteolysis of κ-caseinand subsequent coagulation of milk are of major biological sig-nificance (Mercier et al. 1976, Martin et al. 2011)

A grouping system for mammals based on κ-casein structureand the site of cleavage by chymosin has been suggested(Mercier et al. 1976, Nakhasi et al. 1984). Group I species (cow,goat, sheep and buffalo) have a higher content of dicarboxylicamino acids and low hydrophobicity and carbohydrate contentand κ-casein is cleaved at Phe105–Met106, while Group II species(horse, human, mouse, pig, rat) have a high proline content, lessdicarboxylic amino acids and a much higher hydrophobicityand carbohydrate content and are cleaved at Phe97–Ile98 orPhe105–Leu106. Marsupial κ-casein appears to form a separategroup with a cleavage site different from that in eutherianmammals (Stasiuk et al. 2000). Cleavage of equine milk atPhe97–Ile98, as well as other characteristics of its κ-casein,place the horse in Group II. The divergence between speciesinto Groups I and II could account for differences in the clottingmechanisms of ruminant and non-ruminant milks (Herskovits1966). In addition to the differences in cleavage site, thegrouping system also divides species based on the number ofO-glycosylation sites in κ-caseins. As equine and humanκ-casein are considerably more highly glycosylated than bovineκ-casein and non-glycosylated κ-casein has not been found inequine milk (Egito et al. 2001), equine and human κ-caseinsbelong to the same group. The level of glycosylation does notaffect micelle structure but it does affect the susceptibility of κ-casein to hydrolysis by chymosin, with susceptibility decreasingas the level of glycosylation increases (Doi et al. 1979, Addeoet al. 1984, Van Hooydonk et al. 1984, Vreeman et al. 1986,Zbikowska et al. 1992). Therefore, equine milk probably has adifferent clotting mechanism by chymosin than bovine milk.

Equid Casein Micelles

In the milk of all species studied in sufficient detail, the caseinsexist predominantly as micelles, which are hydrated spheri-

cal structures with dimensions in the sub-micron range. Thedry matter of casein micelles consists predominantly (>90%)of proteins, with small amounts of inorganic matter, collec-tively referred to as micellar calcium phosphate (MCP). Thestructure and sub-structure of bovine casein micelles has beenstudied in detail and reviews include: Holt and Horne (1996),Horne (1998, 2006), De Kruif and Holt (2003), Phadungath(2005), Farrell et al. (2006), Qi (2007), Fox and Brodkorb(2008).

Equine casein micelles are larger than bovine or human mi-celles (Table 26.2) (Welsch et al. 1988, Buchheim et al. 1989)while those of asinine milk are similar in size to bovine micelles(Salimei 2011). Electron microscopy shows that bovine andequine micelles have a similar ‘spongy’ appearance, while hu-man micelles seem to have a much ‘looser’, more open structure(Jasinska and Jaworska 1991). Such a loose open structure mayaffect the susceptibility to hydrolysis by pepsin. Jasinska andJaworska (1991) reported that human micelles are much moresusceptible to pepsin hydrolysis than either equine or bovinemicelles. There are no specific reports on the sub-structure ofequine casein micelles although equine milk does contain ap-proximately 10.1 mmol.L−1 micellar calcium and approximately2.6 mmol.L−1 micellar inorganic phosphate, suggesting a micel-lar calcium:casein ratio of >20:1 which, on a molar basis, far ex-ceeds the calcium-binding capacity of equine casein molecules.Hence, it may be assumed that equine micelles, like bovine ca-sein micelles, contain nanoclusters of calcium phosphate. Sinceequine milk contains little or no κ-casein, unphosphorylated β-casein may play a role in micellar stability (Ochirkhuyag et al.2000, Doreau and Martin-Rosset 2002). A similar conclusionwas reported by Dev et al. (1994) for the stabilisation of humancasein micelles.

Both equine αs1-casein (residues 75–81) and β-casein(residues 23–28) contain a phosphorylation centre, which isrequired for the formation of nanoclusters; furthermore, bothproteins also contain distinct hydrophobic regions throughwhich solvent-mediated protein–protein interactions may oc-cur. Equine αs2-casein may have similar properties to equineαs1-casein, pending further characterisation. The ratio of micel-lar calcium:micellar inorganic phosphate is 2.0 in equine milk,

P1: SFK/UKS P2: SFK



but approximately 3:9 in bovine milk (Holt and Jenness 1984)and might indicate that either a smaller proportion of micellarcalcium is incorporated into nanoclusters in equine milk, or thatequine nanoclusters contain a higher proportion of casein-boundphosphate, which would imply smaller nanoclusters. However,unlike bovine κ-casein, equine κ-casein does not have a dis-tinctly hydrophilic C-terminal domain; thus, it is unclear if thispart of the protein is capable of protruding from the micellarsurface to sterically stabilise the micelles. Furthermore, giventhat the size of casein micelles and the content of κ-casein areinversely related (Yoshikawa et al. 1982, Dalgleish 1998.), alow level of κ-casein would be expected in equine milk com-pared to bovine milk. Further research is required to elucidatethe structure of equine and asinine casein micelles as destabili-sation of the micelles is the basis for the successful conversionof milk into a range of dairy products, for example cheese oryoghurt.

Stability of Equid Casein Micelles

Coagulation of milk occurs when the colloidal stability of thecasein micelles is destroyed and may be desirable or undesir-able. Coagulation is desirable in the manufacture of yoghurtand cheese and is also important from a nutritional point ofview, as clotting of the caseins in the stomach, and the typeand structure of the resultant coagulum strongly affect di-gestibility. In contrast, heat-induced coagulation of casein mi-celles, which can occur at a temperature >120◦C, is undesir-able. In this section, common types of micellar instability aredescribed.

Bovine casein micelles are sterically stabilised by a brush ofpredominantly κ-casein (De Kruif and Zhulina 1996), whichprotrudes from the micelle surface. Coagulation of casein mi-celles can occur only following collapse of the brush, whichoccurs on acidification of milk, that is in the manufacture of yo-ghurt or on removal of the brush that occurs on rennet-inducedcoagulation of milk. The combined process of enzyme- andacid-induced coagulation is likely to contribute to coagulationof casein micelles in the stomach.

Enzymatic Coagulation of Equid Milk

Enzymatic coagulation of milk is the first step in the manufac-ture of most cheese varieties and also plays an important rolein the flocculation of casein micelles in the stomach. For cheesemanufacture, the process involves the addition of a milk-clottingenzyme, for example chymosin, to the milk, followed by incu-bation at a temperature ≥30◦C. During the incubation of bovinemilk with rennet, chymosin hydrolyses the Phe105–Met106 bondof κ-casein, leading to the formation of two fragments, the hy-drophobic N-terminal fragment, f1–105, which remains attachedto the casein micelles and is referred to as para- κ-casein, and thehydrophilic C-terminal fragment, f106–169, which is releasedinto the milk serum and is referred to as the CMP. As a result, themicelles lose steric stabilisation and become susceptible to ag-gregation, particularly in the presence of Ca2+ (for reviews, seeWalstra and Jenness 1984, Wong et al. 1988, Walstra 1990, Foxand McSweeney 1998, Walstra et al. 2006b). Equine κ-casein ishydrolysed slowly by chymosin at the Phe97–Ile98 bond (Kottsand Jenness 1976, Egito et al. 2001), without gel formation,and it appears that either the chymosin-sensitive bond of equineκ-casein is located in the micelle in a manner that renders it inac-cessible by chymosin, or that the equine casein micelle derivescolloidal stability from constituents other than κ-casein. Thehigh degree of glycosylation may also affect the ability of chy-mosin to hydrolyse equine κ-casein. Figure 26.2 illustrates thecoagulation of equine, asinine and bovine milk by calf chymosinat 30◦C. While it is clear that no gel is formed from equine milk,as judged by lack of an increase in storage modulus, G1, asininemilk seems to form a gel, although it is very weak compared tothe gel formed from bovine milk. Further investigation is war-ranted to determine if there are differences in the coagulationproperties of asinine and equine milk.

Acid-induced Coagulation of Equid Milk

When bovine milk is acidified to a pH below 5.0, flocculationof casein micelles occurs, leading ultimately to gel formation.This process is the basis of the manufacture of yoghurt, in whichacidification is induced by the production of lactic acid by lacticacid bacteria and also occurs at the low pH of the stomach (for

0.1

1

10

100

0 20 40 60 80 100 120

0.0001

0.001

0.01

Time (min)

Lo

g G

' (P

a)

Figure 26.2. Rennet-induced coagulation of equine milk (----), asinine milk (---) and bovine milk ( ) at 30◦C.

P1: SFK/UKS P2: SFK


502 Part 4: Milk

1.0000

10.0000

100.0000

0.0001

0.0010

0.0100

0.1000

Lo

g G

' (P

a)11610492806856453624120

Time (min)

Figure 26.3. Coagulation of equine milk (----), asinine milk (---) and bovine milk ( ) acidified with 3% glucono-δ-lactone at 30◦C.

review, see Lucey and Singh 2003). Acid-induced flocculationof bovine casein micelles is believed to result from a reductionin the solvency of the κ-casein brush on the micellar surfacedue to protonation of the negatively charged carboxylic acidgroups of Glu and aspartic acid (Asp). Equine casein micellesare considerably less susceptible to acid-induced flocculation. DiCagno et al. (2004) reported that equine milk acidified at pH 4.2,the point of minimum solubility of equine caseins (Egito et al.2001), had an apparent viscosity only approximately seven timeshigher than that of equine milk at its natural pH (Waelchli et al.1990) and is probably indicative of micellar flocculation ratherthan gelation. By comparison, the viscosity of acidified bovinemilk is approximately 100 times higher than that of bovine milkat natural pH. Differences in acid-induced flocculation betweenequine and bovine casein micelles may be related to differencesin the mechanism by which they are sterically stabilised. Figure26.3 shows the effect of acidification of bovine, equine andasinine milk at 30◦C and 3% glucono-δ-lactone (GDL). Asininemilk appears to form a weak gel when treated with GDL, unlikeequine milk that shows little or no gel formation. Elucidation ofthe mechanism of steric stabilisation of equine casein micellesis likely to shed further light on this subject.

Heat-induced Coagulation of Equine Milk

Although milk, compared to most other foods, is extremely heat-stable, coagulation does occur when heated for a sufficiently longtime at >120◦C. Unconcentrated bovine milk, usually assayedat 140◦C, displays a typical profile, with a heat coagulationtime (HCT) maximum (∼20 minutes) at pH approximately 6.7and a minimum at pH approximately 6.9 (O’Connell and Fox2003). In contrast, the HCT of unconcentrated equine milk at140◦C increases with pH, that is, it has an almost sigmoidalpH-HCT profile (Fig. 26.4), from <2 minutes at pH 6.3–6.9 to>20 minutes at pH 6.9–7.1; a slight maximum is observed at pH7.2. Pre-heating unconcentrated milk shifts the pH-HCT profileand reduces the HCT in the pH region around the maximum,similar to the effect reported for bovine milk (O’Connell andFox 2003). The HCT of concentrated equine milk at 120◦C in-creases up to pH 7.1 but decreases progressively at higher pHvalues. While the profile for concentrated bovine milk is some-

what similar, the maximum HCT occurs at a considerably lowerpH, that is, approximately 6.6. Differences in heat stability be-tween equine and bovine milk may be related to differencesin steric stabilisation of the micelles and, while heat-inducedcomplexation of β-Lg with κ-casein greatly affects the heat sta-bility of bovine milk (O’Connell and Fox 2003), it is unlikelyto do so in equine milk due to lack of a sulphydryl group inequine β-Lg. The lower protein, particularly casein, concentra-tion in equine milk is also likely to contribute to its higher heatstability.

The colloidal stability of equine casein micelles differs con-siderably from that of bovine casein micelles, which may havesignificant implications for the conversion of equine milk intodairy products. On the basis of the evidence outlined, manufac-ture of cheese and yoghurt from equine milk is unlikely to besuccessful using conventional manufacturing protocols.

Stability of Equine Milk to Ethanol

The ethanol stability of bovine milk (for review, see Horne2003), defined as the minimum concentration of added aque-ous ethanol that causes it to coagulate at its natural pH (∼6.7),is 70–75% (added 1:1 to milk), whereas the ethanol stabilityof equine milk (pH ∼7.2) is 40–45% (Uniacke-Lowe, 2011).The high concentration of ionic calcium and low level ofκ-casein in equine milk probably contribute to its low ethanolstability.

Whey Proteins

Similar to bovine milk, the major whey proteins in equine andasinine milk areβ-Lg, α-La, Igs, blood serum albumin (BSA), Lfand Lyz (Bell et al. 1981a, Salimei et al. 2004, Guo et al. 2007).Except for β-Lg, all these proteins are also present in humanmilk. However, the relative amounts of the whey proteins differconsiderably between these milks (Table 26.2). Compared tobovine milk, equine milk contains less β-Lg but more α-Laand Igs. The principal anti-microbial agent in equine milk isLyz and to a lesser extent Lf (which predominates in humanmilk (Table 26.2). Both Lf and Lyz are present at low levelsin bovine milk, in which Igs form the main defense against

P1: SFK/UKS P2: SFK



1200

1400

1600

600

800

1000

0

200

400

7.87.77.67.57.47.37.27.176.96.86.76.66.56.46.3

pH

Hea

t co

agu

lati

on

tim

e (s

)

Figure 26.4. Heat coagulation time-pH profile of raw unconcentrated skimmed equine milk at 140◦C (-----), preheated and unconcentratedmilk ( ) and concentrated milk at 120◦C (---).

microbes (Malacarne et al. 2002). Together, Ig A, G, M, Lfand Lyz provide the neonate with immune and non-immuneprotection against infection (Baldi et al. 2005).

β-Lactoglobulin

β-Lg is the major whey protein in the milk of most ruminants andis also present in the milk of some monogastrics and marsupials,but is absent from the milk of humans, camels, lagomorphs androdents. β-Lg is synthesised in the secretory epithelial cells ofthe mammary gland under the control of prolactin. Although sev-eral biological roles for β-Lg have been proposed, for examplefacilitator of vitamin A (retinol) uptake and an inhibitor, modifieror promoter of enzyme activity, conclusive evidence for a spe-cific biological function of β-Lg is not available (Sawyer 2003,Creamer and Sawyer 2011). β-Lg of all species studied bindsretinol; β-Lg of many species, but not equine or porcine, bindsfatty acids also (Perez et al. 1993). During digestion, milk lipidsare hydrolysed by pre-gastric and pancreatic lipases, greatly in-creasing the amount of free fatty acids that could potentially bindto β-Lg, displacing any bound retinol, and implying that fattyacid metabolism, rather than retinol transport, is the more impor-tant function of β-Lg (Perez and Calvo 1995). Bovine β-Lg isvery resistant to peptic digestion and can cause allergenic reac-tions on consumption. Resistance to digestion is not consistentamong species, with ovine β-Lg being far more digestible thanbovine β-Lg (El-Zahar et al. 2005). The digestibility of equineβ-Lg, which has, to our knowledge, not been studied, warrantsresearch, particularly considering the potential applications ofequine milk as a hypo-allergenic dairy product.

Two isoforms of equine β-Lg have been isolated, β-Lg Iand II, which contain 162 and 163 amino acids, respectively.The extra amino acid in equine β-Lg II is a glycine residueinserted after position 116 of β-Lg I (Halliday et al. 1991).Asinine milk also has two forms of β-Lg, I and II (MW. 18.5and 18.2 kDa, respectively); two variants of β-Lg I, that is, Aand B, are known and four variants of β-Lg II, A, B, C andD (Cunsolo et al. 2007). Godovac-Zimmermann et al. (1985,1988a,b) reported that β-Lg I from asinine milk has 162 aminoacids, similar to equine β-Lg I (from which it differs by 5 aminoacids). β-Lg II in both asinine and equine milks has 163 aminoacids and shows substantial differences between both milks, withonly six clusters of amino acid residues conserved (Godovac-Zimmermann et al. 1990). Criscione et al. (2009) reported theabsence of β-Lg II from more than 23% of Ragusana donkeys inone study.

Bovine β-Lg occurs mainly as two genetic variants, A andB, both of which contain 162 amino acids and differ only atpositions 63 (Asp in variant A, Gly in variant B) and 117(valine (Val) in variant A, alanine (Ala) in variant B); a fur-ther 11, less common, genetic variants of bovine β-Lg havealso been reported (Sawyer 2003). On the basis of its aminoacid sequence, unmodified equine β-Lg I has a molecular massof 18,500 Da and an isoelectric pH of 4.85, whereas equineβ-Lg II, despite having one more amino acid, has a molecu-lar mass of 18,262 Da (ExPASy ProtParam Tool 2009), andan isoelectric pH of 4.71 (Table 26.5). Bovine β-Lg A andB have a molecular mass of 18,367 and 18,281 Da, respec-tively, and an isoelectric pH of 4.76 and 4.83, respectively(Table 26.5). Using the hydropathy scale proposed by Kyte andDoolittle (1982), equine β-Lg I and II have a GRAVY score of

P1: SFK/UKS P2: SFK


504 Part 4: Milk

Table 26.5. Properties of Equine and Bovine β-Lactoglobulin (β-Lg) and Equine Bovine and Human α-Lactalbumin(α-La) and Lactoferrin.

Protein Species Variant

PrimaryAccessionNumbera

AminoAcid

ResiduesMolecularMass (Da)

Isoelectricpoint

GRAVYScoreb

DisulphideBridges

β-Lg Equine I P02758 162 18500.2 4.85 −0.386 2II P07380 163 18261.6 4.71 −0.300 2

Bovine A P02754 162 18367.3 4.76 −0.167 2B P02754 162 18281.2 4.83 −0.162 2

α-La Equine A P08334 123 14223.2 4.95 −0.416 4c

B P08896 123 14251.2 4.95 −0.503 4c

C P08896 123 14249.3 5.11 −0.438 4c

Bovine P00711 123 14186.0 4.80 −0.453 4Human P00709 123 14078.1 4.70 −0.255 4

Lactoferrin Equine O77811 689 75420.4 8.32 −0.376 17Bovine P24627 689 76143.9 8.67 −0.350 16d

Human P02788 691 76165.2 8.47 −0.415 16

Source: Modified from Uniacke et al. 2010.Values were calculated from the amino acid sequences of the mature proteins provided on http://au.expasy.org.aPrimary accession number for the protein in SWISS-PROT database.bGrand average hydropathy (GRAVY) score using the scale of Kyte and Doolittle (1982).cEstimated from structural similarity with bovine and human α-La.dEstimated from structural similarity with human lactoferrin.

−0.386 and −0.300, respectively (Table 26.5). Bovine β-Lg Aand B have a GRAVY score of −0.167 and −0.162, respectively(Table 26.5), and are, therefore, considered to be less hydrophilicthan equine β-Lg I and II. Both equine and bovine β-Lg containtwo intramolecular disulphide bridges, linking Cys66 to Cys160

and Cys106 to Cys119 in equine β-Lg I, Cys66 to Cys161 and Cys106

to Cys120 in equine β-Lg II and Cys66 to Cys160 and Cys106 toCys119 or Cys121 in bovine β-Lg A and B. Bovine β-Lg containsone sulphydryl group at Cys119 or Cys121. Equine β-Lg containsonly four cysteine residues and lacks a sulphydryl group thathas major implications for denaturation and aggregation of theprotein (see later).

At physiological conditions (neutral pH and β-Lg concentra-tion >50 µM), bovine β-Lg occurs predominantly in dimericform and at its isoelectric point (pH 3.7–5.2) the dimers as-sociate into octamers but below pH 3.4 and above pH 8.0 theprotein dissociates into its monomeric form (Gottschalk et al.2003). Equine and asinine β-Lg I exist in the monomeric formonly (Godovac-Zimmermann et al. 1990).

α-Lactalbumin

α-La, a unique milk protein, is homologous with the well-characterised C-type Lyz. It is a calcium metalloprotein, in whichthe Ca2+ plays a crucial role in folding and structure and hasa regulatory function in the synthesis of lactose (Larson 1979,Brew 2003, Neville 2009).

Similar to the α-La of asinine, bovine, caprine, ovine, camelidand human milk, equine α-La contains 123 amino acids (Brew2003). Equine α-La occurs as three genetic variants, A, B

and C, which differ by only a few single amino acid replace-ments (Godovac-Zimmermann et al. 1987). Bovine α-La occursas two, or possibly three, genetic variants (Bell et al. 1981b)and human α-La has two genetic variants, one of which hasbeen identified only recently (Chowanadisai et al. 2005). Theprimary structure of equine, bovine and human α-La differ onlyby a few single amino acid replacements, and the proteins havesimilar properties (Table 26.5). Equine α-La A, B and C have anisoelectric point at pH 4.95, 4.95 and 5.11, respectively, whereasbovine and human α-La have isoelectric point at pH 4.80 and4.70, respectively (Table 26.5). The GRAVY scores of equineand bovine α-La are comparable, whereas that of human α-La isdistinctly higher (Table 26.5), indicating a lower hydrophobicity.The eight-cysteine residues of bovine and human α-La form fourintramolecular disulphide bonds, linking Cys6 to Cys120, Cys28

to Cys111, Cys61 to Cys77 and Cys73 to Cys93. On the basis of thevery high similarity between equine, bovine and human α-Las,as well as the α-La of other species, it is very likely that equineα-La also contains four intramolecular disulphide bridges, in theaforementioned positions. Equine, bovine or human α-La doesnot contain a sulphydryl group.

Three genetic variants of equine α-La have been reportedbut asinine milk has only one (123 amino acid residues, Mwapproximately 14.2 kDa and four disulphide bonds), althoughsome heterogeneity has been shown. Two isoforms, A and B, ofasinine α-La (whose isoelectric points differ by 0.23 units) havebeen reported but subsequent analysis showed that the proteinhas only one form and misidentification in earlier work wasprobably due to differences in calcium binding by asinine α-La(Giuffrida et al. 1992). The primary structure of asinine α-La

P1: SFK/UKS P2: SFK



has been determined and differs from those of equine and bovineproteins with 39 and 40 amino acid substitutions, respectively(Godovac-Zimmermann et al. 1987).

Immunoglobulins

The concentration of whey proteins is significantly elevated inthe colostrum of all ruminants and equids as maternal Igs arepassed from mother to neonate after birth when the small in-testine is capable of absorbing intact proteins. After a few days,the gut ‘closes’ and further significant passage of proteins is pre-vented and within 2–3 days, the serum level of IgG in the neonateis similar to adult levels (Widdowson 1984). In contrast, in uterotransfer of Igs occurs in humans and in some carnivores Igs arepassed to the newborn both before and after birth. The milk ofspecies that provide prenatal passive immunisation tends to haverelatively small differences in protein content between colostrumand mature milk compared to species that depend on post-natalpassage of maternal Igs. In the latter cases, of which all un-gulates are typical, colostrum is rich in Igs and there are largequantitative differences in protein content between colostrumand mature milk (Langer 2009).

Three classes of Igs, which form part of a mammal’s naturaldefense against infection, are commonly found in milk, IgIgG,IgA and IgM; IgG is often sub-divided into two sub-classes, IgG1

and IgG2 (Hurley 2003, Madureira et al. 2007). All monomericIgs consist of a similar basic structure of four polypeptides, twoheavy chains and two light chains, linked by disulphide bridges,yielding a sub-unit with a molecular mass of approximately 160kDa. IgG consists of one sub-unit, while IgA and IgM consist oftwo or five sub-units, with a molecular mass of approximately400 or approximately 1000 kDa, respectively. The relative pro-portions of the Igs in milk differ considerably between species(Table 26.2). IgG is the principal Ig in equine colostrum, butIgA is the principal form in equine milk. In bovine milk andcolostrum, IgG is the principal immunoglobulin, while IgA isthe predominant Ig in human colostrum and milk.

Lactoferrin

Lf is an iron-binding glycoprotein, comprising of a singlepolypeptide chain of MW approximately 78 kDa (Conneely2001). Lf is structurally very similar to transferrin (Tf), a plasmairon transport protein, but has a much higher (∼300-fold) affin-ity for iron (Brock 1997). Lf is not unique to milk although it isespecially abundant in colostrum, with small amounts in tears,saliva and mucus secretions and in the secondary granules ofneutrophils. The expression of Lf in the bovine mammary glandis dependent on prolactin (Green and Pastewka 1978); its con-centration is very high during early pregnancy and involutionand is expressed predominantly in the ductal epithelium closeto the teat (Molenaar et al. 1996). Human, equine, asinine andbovine milk contain 1.65 , 0.58 , 0.37 and 0.1 g Lf/kg, respec-tively (Table 26.2). The concentration of Lf in asinine milk,which comprises approximately 4% of total whey protein, issignificantly lower than in equine milk (Table 26.2).

Shimazaki et al. (1994) purified Lf from equine milk andcompared its iron-binding ability with that of human and bovineLfs and with bovine Tf. The iron-binding capacity of equine Lfis similar to that of human Lf but higher than that of bovine Lfand Tf. Various biological functions have been attributed toLf but the exact role of Lf in iron-binding in milk is unknownand there is no relationship between the concentrations of Lfand Tf and the concentration of iron in milk (human milk is veryrich in Lf but low in iron) (Masson and Heremans 1971).

Lf is a bioactive protein with nutritional and health-promotingproperties (Baldi et al. 2005). Bacterial growth is inhibited by itsability to sequester iron and also to permeabilise bacterial cellwalls by binding to lipopolysaccharides through its N-terminus.Lf can inhibit viral infection by binding tightly to the envelopeproteins of viruses and is also thought to stimulate the establish-ment of a beneficial microflora in the gastrointestinal tract (Baldiet al. 2005). Ellison and Giehl (1991) suggested that Lf andLyz work synergistically to effectively eliminate Gram-negativebacteria; Lf binds oligosaccharides (OSs) in the outer bacterialmembrane, thereby opening ‘pores’ for Lyz to hydrolyse glyco-sidic linkages in the interior of the peptidoglycan matrix. Thissynergistic process leads to inactivation of both Gram-negative,for example E. coli (Rainhard 1986) and Gram-positive bacte-ria, for example Staphylococcus epidermidis (Leitch and Will-cox 1999) bacteria. Furthermore, a proteolytic digestion productof bovine and human Lf, lactoferricin, has bactericidal activity(Bellamy et al. 1992). Bovine and human Lf are reported tohave antiviral activity and a role as a growth factor (Lonnerdal2003). The specific biological function of equine Lf has not beenstudied, but is likely to be similar to that of bovine and human Lf.

Equine Lf contains 689 amino acid residues, which is sim-ilar to bovine Lf and two more than human Lf (Table 26.5).Compared to most other milk proteins, Lf has a high isoelec-tric point, that is, at pH 8.32, 8.67 or 8.47 for equine, bovineor human Lf (Table 26.5). As a result, the protein is positivelycharged at the pH of milk and may associate with negativelycharged proteins via electrostatic interactions. GRAVY scoresare comparable for equine, bovine and human Lf (Table 26.5).Equine and human Lf contain 17 and 16 intra-molecular disul-phide bonds, respectively (Table 26.5). On the basis of structuralsimilarities with human Lf, it has been assumed that bovineLf contains 16 intra-molecular disulphide bonds (Table 26.5).The iron-binding capacity of equine, bovine and human Lfs areequivalent, although the pH-dependence of the iron-binding ca-pacity of bovine Lf differs from that of equine and human Lf(Shimazaki et al. 1994).

All Lfs studied to date are glycosylated, but the location andnumber of potential glycosylation sites, as well as the numberof sites actually glycosylated, vary. In bovine Lf, four out of fivepotential glycosylation sites, that is, Asn223, Asn368, Asn476 andAsn545, are glycosylated (Moore et al. 1997), whereas in hu-man Lf, two of three potential glycosylation sites, that is, Asn137

and Asn478, are glycosylated (Haridas et al. 1995). Glycosyla-tion of equine Lf has not been studied, but using the consensussequence, Asn-Xaa-Ser/Thr (where Xaa is not Pro), for glyco-sylation, three potential glycosylation sites are likely in equineLf, that is Asn137, Asn281 and Asn476.

P1: SFK/UKS P2: SFK


506 Part 4: Milk

Whey Protein Denaturation

Whey proteins are susceptible to heat-induced denaturation. Thethermal stability of equine Lf and BSA is comparable to that ofthe bovine proteins but equine β-Lg and α-La are more heatstable than the bovine proteins (Bonomi et al. 1994). Equineβ-Lg is more thermally stable than equine α-La, which is dif-ferent from bovine milk, where α-La is the more thermallystable (Civardi et al. 2007). The high thermal stability of equineβ-Lg may be related to its lack of a sulphydryl group. Thermaldenaturation of bovine β-Lg is a two-stage process, unfoldingof the polypeptide chain and exposure of the sulphydryl group,followed by self-association or interaction with other proteinsvia sulphydryl–disulphide interchange (Sawyer 2003). Owing tothe lack of a sulphydryl group, equine β-Lg cannot undergo thesecond denaturation step and therefore its structure may refoldon cooling. Denaturation of α-La is commonly a result of com-plex formation with β-Lg via sulphydryl–disulphide interchangeand its higher thermal stability may therefore be a result of dif-ferences in its environment, rather than its molecular structure.Recent research suggests that equine α-La and β-Lg are alsoless susceptible to denaturation than their bovine counterpartsunder high pressure.

DIGESTIBILITY OF EQUID MILKMilk is highly digestible and, because it is liquid, the gastroin-testinal tract of mammals has mechanisms for delaying its pas-sage; coagulation of milk in the stomach delays the degradationof proteins and improves their assimilation by the body. Caseinsare precipitated by gastric acid and enzymes, forming a clotin the stomach that entraps fat. The hardness of this clot de-pends on the casein content of the milk; high casein-containingmilks will produce firm clots. Generally, species that nurse theiryoung at frequent intervals, for example equids and humans,tend to produce dilute milk in which <60% of total protein iscasein and which form a soft clot, whereas those that nurse infre-quently, for example cattle and sheep, produce milk that is highin fat and casein and has much longer gastric retention (Jenness1986). Degradation of casein in the gastrointestinal tract is slowbut extensive and while β-Lg is relatively resistant to gastricproteolysis, α-La is readily hydrolysed when the gastric pH isapproximately 3.5 (Savalle et al. 1988).

The physico-chemical differences between human and bovinecaseins result in the formation of different types of curd in thestomach (Hambræus 1982) and because the protein profile ofequine milk is quite similar to that of human milk, equine milkmay be more suitable in human nutrition than bovine milk.Turner (1945) compared the digestibility of equine, human andbovine milk based on the average percentage conversion ofacid-insoluble protein to acid-soluble protein during digestion.Equine and human milk have a much lower buffering capacitythan bovine milk and, while equine milk is very digestible, it isslightly less than human milk but significantly better than bovinemilk. Turner (1945) concluded that both equine and human milkform soft curds in the stomach that pass through the digestivetract more quickly than bovine milk curd. Kalliala et al. (1951)

also reported that the overall digestibility of equine and humanmilk (by in vitro experiments) appeared to be quite similar andboth were easier to digest than bovine milk. Human milk formsfine, soft flocs in the stomach with an evacuation time of 2–2.5hours, whereas bovine milk forms compact hard curds with adigestion time of 3–5 hours.

TOTAL AMINO ACIDSGuo et al. (2007) investigated the total amino acid compositionof asinine milk and expressed the results both in grams of in-dividual amino acids per 100 grams of milk and per 100 gramsof protein and compared values to those for equine, bovine andhuman milk (Table 26.6). Results expressed per 100 g of milkdemonstrated differences related, most likely, to differences inthe total protein content between the milks, but when expressedas g/100 g protein, the differences were not so apparent. It hasbeen reported that glycine is exceptionally high in equine casein(Lauer and Baker 1977) and other studies have reported that themean values of peptide-bound amino acids in equid milks aregenerally higher than those in camel and buffalo milks and mayindicate that equid milks are more suitable for human consump-tion than other milks studied to date. Asinine milk has noticeablyhigher levels of serine, glutamate, arginine and valine and muchless cystine and the percentage of seven of the eight essentialamino acids (isoleucine, leucine, lysine, methionine, phenylala-nine, threonine, tyrosine, valine) is also higher than those ofequine and bovine milk (Guo et al. 2007). In equine, bovineand human milk cystine, glycine, serine, threonine and alaninedecrease as lactation progresses while glutamate, methionine,isoleucine and lysine tend to increase (Davis et al. 1994).

NON-PROTEIN NITROGENThe non-protein nitrogen (NPN) of milk consists primarily ofurea, peptides, amino acids and ammonia. NPN constitutes10–15% of the total nitrogen in mature equine milk which isintermediate between the values for human milk and ruminantmilk, 25% and 5%, respectively (Hambræus 1984, Oftedal et al.1983, Atkinson et al. 1989, Walstra et al. 2006b). In equinemilk, NPN increases from <2% of total nitrogen at parturi-tion to >10% after 2 weeks (Zicker and Lonnerdal 1994). Thecomponents of the NPN in human and bovine milk have beencharacterised (see Atkinson et al. 1989, Atkinson and Lonnerdal1995, Rudloff and Kunz 1997, Carratu et al. 2003), but the NPNof equine milk has not been studied in detail.

Up to 50% of the NPN of human milk is urea and free aminoacids, the exact function of which is, as yet, unknown. Asininemilk has a significantly higher level of NPN than equine milkand is close to that of human milk (Table 26.2). Equine andasinine milk have similar urea levels.

Free Amino Acids

The free amino acid content of equine, bovine and human milkare 1960, 578 and 3019 µmol.L−1, respectively (Rassin et al.1978, Agostini et al. 2000) (Table 26.7). Glutamine, glutamate,

P1: SFK/UKS P2: SFK



Table 26.6. Amino Acid Composition of Asinine and Equine MilkExpressed as g amino acid per 100 g Protein, with Comparative Datafor Bovine and Human Milk

Amino Acid Asinine Equine Bovine Human

Aspartic acid 8.9 10.4 7.8 8.3Serine 6.2 6.2 4.8 5.1Glutamic acid 22.8 20.1 23.2 17.8Glycine 1.2 1.9 1.8 2.6Histidine 2.3 2.4 3.0 2.3Arginine 4.6 5.2 3.3 4.0Threonine 3.6 4.3 4.5 4.6Alanine 3.5 3.2 3.0 4.0Proline 8.8 8.4 9.6 8.6Cystine 0.4 0.6 0.6 1.7Tyrosine 3.7 4.3 4.5 4.7Valine 6.5 4.1 4.8 6.0Methionine 1.8 1.5 1.8 1.8Lysine 7.3 8.0 8.1 6.2Isoleucine 5.5 3.8 4.2 5.8Leucine 8.6 9.7 8.7 10.1Phenylalanine 4.3 4.7 4.8 4.4Tryptophan – 1.2 1.5 1.8Essential amino acids 38.2 36.7 37.5 40.7

Source: Modified from Guo et al. 2007.

Table 26.7. Free Amino Acids (µM.L−1) of Equine, Bovine andHuman Milk

Amino Acid Equinea Bovinea Humanb

Alanine 105 30.0 227.5Arginine 14.0 10.0 35.4Aspartic acid 40.0 15.0 183.2Cystine 2.0 21.0 56.0Glutamic acid 568.0 117.0 1184.1Glutamine 485.0 12.0 284.8Glycine 100.0 88.0 124.6Histidine 46.0 9.0 7.7Isoleucine 8.0 3.0 33.4Leucine 16.0 3.0 55.6Lysine 26.0 15.0 39.0Methionine ∼0 ∼0 8.8Phenylalanine 5.0 3.0 23.6Proline 1.61 – 64.3Serine 175 23.0 273.7Taurine 32.0 13.0 301.1Threonine 137.0 16.0 97.6Tyrosine 3.0 0.3 2.5Valine 45.0 5.0 72.7Total ∼1960.0 578.0 3019.7

aRassin et al. 1978.bAgostini et al. 2000.

P1: SFK/UKS P2: SFK


508 Part 4: Milk

glycine, alanine and serine are the most abundant free aminoacids in equine, bovine and human milk, and taurine also is ex-ceptionally high in human milk (Rassin et al. 1978, Sarwar et al.1998, Carratu et al. 2003). Taurine is an essential metabolitefor the human infant and may be involved in the structure andfunction of retinal photoreceptors (Agostini et al. 2000). Com-pared to bovine milk, equine milk has an appreciable amountof taurine although it is ten times less than that of human milk(Table 26.7). In contrast to total amino acid composition, whichis essentially similar in equine, bovine and human milks, freeamino acids show a pattern characteristic of each species (Table26.7), which may be important for early post-natal developmentin different animals. Free amino acids are more easily absorbedthan protein-derived amino acids and glutamic acid and glu-tamine, which comprise >50% of the total free amino acids ofhuman milk, are a source of α-ketoglutaric acid for the citricacid cycle and also act as neurotransmitters in the brain (Levy1998, Agostini et al. 2000).

Bioactive Peptides

Both caseins and whey proteins are believed to contribute tohuman health through latent biological activity produced en-zymatically during digestion, fermentation with specific startercultures or enzymatic hydrolysis by microorganisms, resultingin the formation of bioactive peptides. These peptides are im-portant for their physiological roles, their opioid-like features,as well as their immunostimulating and anti-hypertensive activi-ties and their ability to enhance Ca2+ absorption and are releasedor activated during gastrointestinal digestion. Several peptidesgenerated by the hydrolysis of milk proteins are known to reg-ulate the overall immune function of the neonate (Baldi et al.2005). A detailed discussion on bioactive peptides in milk is out-side the scope of this chapter, for reviews, see Donnet-Hugheset al. (2000), Shah (2000), Malkoski et al. (2001), Fitzgeraldand Meisel (2003), Silva and Malcata (2005), Fitzgerald andMurray (2006), Lopez-Fandino et al. (2006), Michaelidou andSteijns (2006), Thoma-Worringer et al. (2006) and Phelan et al.(2009).

Research on the bioactive peptides derived from equid milkis very limited. Peptides from the hydrolysis of equine β-casein may have a positive action on human health (Doreauand Martin-Rosset 2002). Chen et al. (2010) reported the pres-ence of four novel angiotensin-converting enzyme-inhibitorypeptides in koumiss which may enhance the beneficial effectsof koumiss on cardiovascular health. Peptides with trophic orprotective activity have been identified in asinine milk (Salimei2011).

Hormones and Growth Factors

Leptin is a protein hormone of approximately 16kDa that hasbeen discovered recently in human milk and plays a key rolein the regulation of energy intake and energy expenditure, aswell as functioning in mammary cell proliferation, differentia-tion and apoptosis. Human-like leptin has been isolated fromasinine milk at a level of 3.2–5.4 ng.mL−1 which is similar to

levels reported for other mammals and showed little variationthroughout lactation (Salimei et al. 2002).

Levels of the bioactive peptides, ghrelin and insulin growthfactor I, which play a direct role in metabolism, body composi-tion and food intake, have also been reported for asinine milk at4.5 pg.mL−1 and 11.5 ng.mL−1, respectively, similar to levelsin human milk (Salimei 2011).

Amyloid A

Amyloid A3 (AA3) is a protein produced in the mammary glandand is encoded by a separate gene from that for serum amyloidA (serum AA) (Duggan et al. 2008). AA3 is believed to pre-vent attachment of pathogenic bacteria to the intestinal cell wall(Mack et al. 2003) and may prevent necrotising enterocolitisin human infants (Larson et al. 2003). McDonald et al. (2001)demonstrated the presence of AA3 in the colostrum of cows,ewes, sows and horses. Bovine colostrum has a high concentra-tion of AA3 but by approximately 3 days post-partum the levelsdecline. In bovine milk, the presence of serum AA in milk is anindicator of mastitic infection (Kaneko et al. 2004, Winter et al.2006). In equine colostrum, the concentration of AA3 is consid-erably lower than in milk and consequently may play a crucialrole in intestinal cell protection in the foal especially after gutclosure (Duggan et al. 2008).

INDIGENOUS ENZYMESMilk contains many indigenous enzymes that originate fromthe mammal’s blood plasma, leucocytes (somatic cells), or cy-toplasm of the secretory cells and the milk fat globule mem-brane (MFGM)(Fox and Kelly 2006). The indigenous enzymesin bovine and human milks have been studied extensively butthe enzymes in the milk of other species have been studied onlysporadically. Equine milk probably contains all the enzymes thathave been identified in bovine milk but relatively few studieshave been reported.

Lysozyme

Lyz (EC 3.1.2.17) occurs at high levels in equine, asinine andhuman milk (Table 26.2). Human, equine and asinine milk con-tain 3000, 6000 and >6000 times more Lyz, respectively, thanbovine milk (Salimei et al. 2004, Guo et al. 2007) with levelsas high as 0.4 g.100g−1 for Martina Franca donkeys (Coppolaet al. 2002) although 0.1 g/100 g is reported and more commonlyfound in asinine milk (Vincenzetti et al. 2008). The concentra-tion of Lyz in human milk increases strongly after the secondmonth of lactation, suggesting that Lyz and Lf play major rolesin fighting infection in breast-fed infants during late lactation,and protect the mammary gland (Montagne et al. 1998).

Equine milk Lyz is more stable to denaturation than humanLyz during pasteurisation at 62◦C for 30 minutes but at 71◦C for2 minutes or 82◦C for 15 seconds, the inactivation of both weresimilar (Jauregui-Adell 1975). It has been suggested, but re-search is scarce, that while the composition of breast milk varies

P1: SFK/UKS P2: SFK



widely between well-nourished and poorly nourished mothers,the amount of Lyz is conserved.

Lyz found in egg white, tears and saliva do not generally bindcalcium but equine and canine milk Lyz do and this is believedto enhance the stability and activity of the enzyme (Nitta et al.1987). The binding of a Ca2+ by Lyz is considered to be anevolutionary linkage between non-Ca2+-binding Lyzs and α-La(Tada et al. 2002, Chowdhury et al. 2004). The conformationof the calcium-binding loop of equine Lyz is similar to that ofα-La (Tsuge et al. 1992, Tada et al. 2002) and both equine Lyzand α-La form stable, partially folded, ‘molten globules’ undervarious denaturing conditions (Koshiba et al. 2001,) with thatof equine Lyz being considerably more stable than α-La (Lyster1992, Morozova-Roche 2007). The molten state of canine Lyz issignificantly more stable than that of equine Lyz (Koshiba et al.2000, Spencer et al. 1999). Equine milk Lyz is very resistant toacid (Jauregui-Adell 1975) and proteolysis (Kuroki et al. 1989),and may reach the gut relatively intact.

Asinine Lyz contains 129 amino acids, is a C-type Lyz,binds calcium strongly and has 51% homology to human Lyz(Godovac-Zimmermann et al. 1988b). Two genetic variants ofLyz, A and B, have been reported in asinine milk (Herrouin et al.2000) but only one is found in equine milk. Asinine Lyz is re-markably heat stable and requires 121◦C for 10 minutes for inac-tivation. The Lyz content of equid milks is one of the main attrac-tions for use of these milks in cosmetology as it is reputed to havea smoothing effect on the skin and may reduce scalp inflamma-tion when incorporated into shampoo. Equid milk has very goodantibacterial activity, presumably due to its high level of Lyz.

Other Indigenous Enzymes in Equine Milk

Lactoperoxidase, catalase, amylase, proteinase (plasmin), li-pase, lactate dehydrogenase and malate dehydrogenase havebeen reported in equine milk. Bovine milk is a rich source ofxanthine oxidoreductase (XOR) but the milk of other species forwhich data are available have much lower XOR activity, becausein non-bovine species, most (up to 98% in human milk) of theenzyme molecules lack Mo and are inactive. XOR has not beenreported in equine milk, which is unusual considering the role ofXOR in the excretion of fat globules from the secretory cells andalso considering that equine milk contains quite a high level ofmolybdenum (Mo), which presumably is present exclusively inXOR. Chilliard and Doreau (1985) characterised the lipoproteinlipase activity of equine milk and reported that the milk has highlipolytic activity, comparable to that in bovine milk and higherthat in caprine milk. There are no reports on hydrolytic rancidityin equine milk, which is potentially a serious problem in equinemilk products and warrants investigation.

Plasmin, a serine proteinases, is one of a number of proteolyticenzymes in milk. Visser et al. (1982) and Egito et al. (2002) re-ported γ -caseins in equine milk and, it is therefore assumed, thatequine milk contains plasmin. Humbert et al. (2005) reportedthat equine milk contains five times more plasmin activity thanbovine milk and 90% of total potential plasmin activity was plas-min, with 10% as plasminogen; the plasmin:plasminogen ratioin bovine and human milk is 18:82 and 28:72, respectively.

Alkaline phosphatase (ALP) is regarded as the most importantindigenous enzyme of bovine milk because ALP activity is usedas the index of the efficiency of high-temperature short-timepasteurisation. About 40% of ALP activity in bovine milk isassociated with the MFGM. Equine milk has 35–350 times lessALP activity than bovine milk and there are no reports on ALPin the equine MFGM. Because of the low level of ALP in equinemilk it has been suggested that it is not suitable as an indicator ofpasteurisation efficiency of equine milk (Marchand et al. 2009)although one would expect that once the exact initial concentra-tion of ALP is known, the use of a larger sample size or a longerincubation period would overcome the low level of enzyme.

CARBOHYDRATESLactose and Glucose

The chemistry, properties and applications of lactose are de-scribed in Chapter 24 and have been reviewed extensivelyelsewhere for example Fox (1985, 1997) and McSweeney andFox (2009) and will not be considered here. The concentrationof lactose in asinine milk is high (51–72.5 g.kg−1), probablymarginally higher than equine milk (approximately 64 g.kg−1)(Table 26.2), which is similar to the level in human milk and sig-nificantly higher than that in bovine milk. As an energy source,lactose is far less metabolically complicated than lipids but thelatter provides significantly more energy per unit mass. As wellas being a major energy source for the neonate, lactose affectsbone mineralisation during the first few months post-partum asit stimulates the intestinal absorption of calcium (Schaafsma2003). Equine milk contains a significant concentration of glu-cose, approximately 50 mg/L in colostrum which increases toapproximately 150 mg/L 10 days post-partum and then de-creases gradually to approximately 120 mg/L (Enbergs et al.1999). Although the lactose content of equid milks is high, thephysico-chemical properties of lactose that cause problems inthe processing of bovine milk are of no consequence for equidmilks, which are consumed either fresh or fermented. In Mongo-lia, where approximately 88% of the population is lactose intol-erant (Yongfa et al. 1984), lactose intolerance is not a problemwith fermented equine milk, koumiss, as approximately 30% oflactose is converted to lactic acid, ethanol and carbon dioxideduring fermentation.

Oligosaccharides

The milk of all species examined contains OSs but the concen-tration varies markedly (see Urashima et al. 2009). The OSsin milk contain 3 to 10 monosaccharides and may be linear orbranched; they contain lactose at the reducing end and also con-tain fucose, galactosamine and N-acetylneuraminic acid. Thehighest levels are in the milk of monotremes, marsupials, ma-rine mammals, humans, elephants and bears. OSs are the thirdmost abundant constituent of human milk that has an exception-ally high content (approximately 20 g.L−1 in colostrum, whichdecreases to 5–10 g.L−1 in milk) and structural diversity ofOSs (>200 molecular species), which have a range of functions,

P1: SFK/UKS P2: SFK


510 Part 4: Milk

Table 26.8. Principal Oligosaccharides of Equine Colostrum

Oligosaccharide (mg/L)

AcidicNeu5Ac(α2–3) Gal(β1—4)Glc N/aGal(β1–4)GlcNAcα1-diphosphate (N-acetyllactosamine-α1-phospahte) N/a

NeutralGal(β1–3)Gal(β1–4)Glc (β3’-galactosyllactose) 7.8Gal(β1–6)Gal(β1–4)Glc (β6’-galactosyllactose) 4.8Gal(β1–4)GlcNAc(β1–3)Gal(β1–4)Glc (lacto-N-neotetraose) N/aGal(β1–4)GlcNAc(β1–6)Gal(β1–4)Glc (iso-lacto-N-neotetraose) 0.5Gal(β1–4)GlcNAc(β1–6)[Gal(β1–3)]Gal(β1–4)Glc (lacto-N-novopentanose 1) 1.1Gal(β1–4)GlcNAc(β1–6)[Gal(β1–4)GlcNAc(β1–3)]Gal(β1–4)Glc (lacto-N-neohexaose) 1.1Gal(β1–4)GlcNAc-1-phosphate (N-acetyllactosamine-1–0-phosphate) N/aNeu5Ac(α2–3)Gal(β1–4)Glc (3’-N-acetylneuraminyllactose) N/a

Source: From Urashima et al. 1989, 2001, Nakamura et al. 2001.Gal, d-galactose; Glc, d-glucose; GlcNAc, N-acetylglucosamine; Neu5A, N-acetylneuraminic acid; N/a, not available.

including as important components of our immune system and asprebiotics to promote a healthy gut microflora (Donovan 2009).Bovine, ovine, caprine and equine milk contain relatively lowlevels of OSs, which have been characterised (see Urashima et al.2001). The OSs identified in equine colostrum are summarisedin Table 26.8. The OSs in mature equine milk have not been re-ported but it can be assumed that the level is considerably lowerthan in colostrum which has approximately 18.6 g/L (Nakamuraet al. 2001). The neutral OSs, lacto-N-neotetraose and lacto-N-neohexaose, in equine colostrum are also abundant in humanmilk, while iso-lacto-N-neotetraose and lacto-N-novopentanose1 are not, but have been identified in bovine colostrum; the latterhas been identified also in the milk of the Tammar wallaby andbrown capuchin monkey (Urashima et al. 2009).

LIPIDSMilk fat is important for the provision of energy to the new-born as well as being the vehicle for fat-soluble vitamins andessential fatty acids. From a practical point of view, milk lipidsare important as they confer distinctive nutritional, textural andorganoleptic properties on dairy products. Dietary compositionis considered one of the major determinants of the fatty acidcomposition of equid milk and non-dietary factors such as stageof lactation, age and parity of the mare play minor roles.

Triglycerides (TGs) represent approximately 80–85% of thelipids in equine and asinine milk, while approximately 9.5% arefree fatty acids (FFAs) and approximately 5–10% are phospho-lipids (Jahreis et al. 1999). In contrast, approximately 97–98%of the lipids in bovine and human milk are TGs, with low lev-els of phospholipids and free fatty acids, 1.3 and 1.5 g.100g−1,respectively. The high level of free fatty acids in equid milkimplies that rancidity is a problem with these milks and is dealtwith in Section ‘Stability of Equine Milk Fat’.

The relatively high content of phospholipids in equid milkis thought to contribute to its buffering properties. TGs, theprimary transport and storage form of lipids, are synthesised

in the mammary gland from fatty acids that originate fromthree sources: de novo synthesis (C8:0, C10:0 and C12:0), di-rect uptake from the blood (>14 carbons) and modificationof fatty acids in the mammary gland by desaturation and/orelongation. Circulating fatty acids in the blood may origi-nate from dietary fat or from lipids mobilised from body fatstores. The principal phospholipids of equid milk are phos-phatidylcholine (19%), phosphatidylethanolamine (31%), phos-phatidylserine (16%) and sphingomylin (34%); the correspond-ing values for bovine milk are 35%, 32%, 3% and 25% and forhuman milk are 28, 20, 8 and 39%, respectively. The high levelof FFAs in equid milk implies considerable lipolysis but thishas not been suggested; if lipolysis was responsible for the highlevel of FFAs, they should be accompanied by high levels ofmono- and di-glycerides but these are reported to be quite lowat approximately 1.8% of total lipids.

Table 26.9 shows the monounsaturated fatty acids (MUFA)and polyunsaturated fatty acids (PUFA) in the milk fat ofsome ruminant and non-ruminant species. The milk fat of non-ruminants contains substantially higher levels of PUFAs thanruminant milks due to the lack of biohydrogenation of fattyacids in the former, and for the two-equid species shown, thehorse has considerably more PUFAs in its milk than the donkey.Saturated fatty acids are the dominant class in asinine milk andlevels are significantly higher than those in equine or humanmilk (Table 26.9).

Fatty Acids in Equid Milks

The fatty acid profile of equid milk (Table 26.10) differs fromthat of bovine and human milk fat in a number of respects.Like human, and unlike bovine milk, equid milk is characterisedby low proportions of saturated fatty acids with low or highernumbers of carbons, that is, C4:0, C6:0, C16:0 and C18:0 (Pikuland Wojtowski 2008). Butyric acid (C4:0) is present at highlevels in bovine and other ruminant milk fats, produced from

P1: SFK/UKS P2: SFK



Table 26.9. Monounsaturated and Polyunsaturated Fatty Acids(Percent of Total Fatty Acids ± Standard Deviations) in the MilkFat of Some Ruminants and Non-Ruminants

MUFAs PUFAs CLA

Non-ruminants

Equine 20.70 36.80 0.09Asinine 15.30 16.00 –Porcine 51.80 12.40 0.23Human 33.20 12.50 0.39

Ruminants

Caprine 26.90 2.58 0.65Bovine 23.20 2.42 1.01Ovine 23.00 3.85 1.08

Data from Jahreis et al. 1999, Salimei et al. 2004.MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids; CLA,conjugated linoleic acid.

3-hydroxybutanoic acid, which is synthesised by bacteria in therumen (Pikul and Wojtowski 2008). Caprylic acid, C8:0, is veryhigh in equid milk compared to the level in human and bovinemilk (Table 26.10). Levels of middle chain-length FAs, espe-cially C10:0 and C12:0, are high in equid milk (20–35% of all FAscontain <16 C) and in all non-ruminant herbivores, suggestingthat they arise from the use of glucose as the principal precursorfor fatty acid synthesis (Palmquist 2006). Mammalian de novofatty acid synthesis requires a carbon source (acetyl-CoA) andreducing equivalents in the form of NADPH + H+. In rumi-nants, acetate and β-hydroxybutyrate are the primary sourcesof carbon while glucose and acetate are the primary sourcesof reducing equivalents; in non-ruminants, for example equidspecies, glucose is the primary source of both carbon and re-ducing equivalents and also supplies some of the glycerol formilk TGs (Dils 1986). When horses were infused with eitherglucose or acetate and palmitate, C12:0 and C14:0 were formedexclusively from acetate, as in ruminants, and C16:0 was formed,partly from acetate and partly from palmitate; unlike ruminants,44% of C18:0 and 7% of C18:1 are formed from acetate in the horse(Palmquist 2006). If this is so, acetate and 3-hydroxybutyrate arepresumably produced by bacterial fermentation in the lower in-testine of the horse and why 3-hydroxybutyrate is not convertedto butanoic acid, as in ruminants, is unclear. Fatty acids fromC6:0 to C16:0 are released from the fatty acid synthesis complexby acyl-specific thioesterases; presumably, the middle chain-length-specific thioesterases are particularly active in equids;investigation of this possibility is warranted.

Equine milk-fat contains a relatively high level of C16:1

(2–10%, w/w) and C18:1, reflecting high �-9 desaturase activity.Equid milk fats contain a very high level of n-3-octadecatrienoicacid (linolenic acid), which reflects the high level of PUFAs inthe diet and the lack of biohydrogenation, as occurs with rumi-nants. In the rumen, extensive hydrogenation of double bonds

occurs and most fatty acids taken up from the intestinal tractare saturated. The large intestine of equids shows significantdifferences in the relative rate of transport of volatile fatty acidscompared to ruminants and de novo synthesis of C18:0 fatty acidsoccurs with a further high proportion of C6:0 to C14:0 carbon fattyacids and some C16:0 arising from products of their large bowelfermentation.

Equine and asinine milks have similar fatty acid profiles al-though the former has a higher content of monosaturated fattyacids (Tables 26.9 and 26.10). Both equine and asinine milk havecharacteristic low levels of stearic acid, and oleic acid is excep-tionally low in asinine milk. Asinine and zebra milk fat contain ahigh level of PUFAs, although considerably lower than in equinemilk. The well-balanced ratio of n-6: n-3 of 1.17:1 in asinine milkcompared to 3.14:1 in equine milk makes it an interesting prod-uct for human nutrition. n-6 and n-3 fatty acids are essential inhuman metabolism as components of membrane phospholipids,precursors of eicosanoids, ligands for membrane receptors andtranscription factors that regulate gene expression. The impor-tance of n-6 C18:2, linoleic acid (LA), has been known for manyyears but the significance of n-3 C18:3, α-linolenic (ALA) was notrecognised until the late 1980s and has since been identified as akey component in the diet for the prevention of atopic dermatitis(Horrobin 2000). LA and ALA are not inter-convertible but arethe parent acids of the n-6 and n-3 series of long chain (LC)polyunsaturated fatty acids, respectively (e.g., n-6 C20:4, arachi-donic acid (AA); n-3 C20:5, eicosapentaenoic acid (EPA) and n-3C22:6, docosahexaenoic acid (DHA)) which are components ofcellular membranes and precursors of other essential metabolitessuch as prostaglandins and prostacyclins (Cuthbertson 1999, In-nis 2007). DHA and AA are now recognised as being crucial fornormal neurological development (Carlson 2001). Humans haveevolved on a diet with a ratio of n-6 to n-3 fatty acids of approx-imately 1:1 but Western diets nowadays have a ratio of 15:1 to

P1: SFK/UKS P2: SFK


512 Part 4: Milk

Table 26.10. Typical Fatty Acid (Percent of Total Fatty Acids) Composition in the Milk of Equid Species; Bovine andHuman Milk are Included for Comparison

Fatty Acid Common Name Equine Asinine Zebra Bovine Human

Saturates

C4:0 Butyric 0.09 0.60 3.90 0.19C6:0 Caproic 0.24 1.22 2.50 0.15C8:0 Caprylic 3.15 12.80 8.20 1.50 0.46C10:0 Capric 6.48 18.65 15.30 3.20 1.03C12:0 Lauric 6.65 10.67 9.20 3.60 4.40C13i:0 0.22C13:0 0.17 3.92 0.19 0.06C14i:0 0.12 0.04C14:0 Myristic 7.04 5.77 6.50 11.10 6.27C15a:0 0.21C15 i:0 0.16 0.07C15:0 Pentadecanoic 0.39 0.32 0.50 1.20 0.43C16i:0 0.12 0.17C16:0 Palmitic 20.43 11.47 13.30 27.90 22.00C17i:0 0.31 0.20 0.23C17:0 Margaric 0.38 0.22 0.60 0.58C18i:0 0.11C18:0 Stearic 1.18 1.12 2.00 12.20 8.06C20:0 0.12 0.35 0.44C21:0 0.04 0.13C22:0 0.05 0.20 0.12C24:0 0.14 0.25Total 46.67 67.66 55.00 68.62 45.33

Monounsaurates

C10:1 1.46 2.20 0.15C12:1 0.20 0.25 0.06C14:1c n-5 Myristoleic 0.52 0.22 0.80 0.41C14:1t n-5 0.07C15:1 Pentadecanoic 0.22 0.30 0.11C16:1c n-7 Palmitoleic 5.68 2.37 4.20 1.50 3.29C16:1c n-9 0.56C16:1t n-7 0.36C17:1 Heptadecanoic 0.62 0.27 0.36 0.37C18:1c n-9 Oleic 20.26 9.65 20.40 17.20 31.30C18:1c n-11 1.31C18:1t n-9 2.67C18:1t n-11 Vaccenic 3.90C20:1 n-9 0.40 0.32 0.67C20:1 n-11 0.35C22:1 n-9 0.06 0.08C24:1 n-9 0.12Total 31.23 15.31 24.65 24.65 39.45

Polyunsaturates

n-6 Series

C18:2cc Linoleic 10.10 8.15 16.04 1.40 10.85C18:2 conj. Conjugated linoleic 0.07 1.10

(Continued)

P1: SFK/UKS P2: SFK



Table 26.10. (Continued)

Fatty Acid Common Name Equine Asinine Zebra Bovine Human

C18:2tt 0.46C18:2ct 0.69C18:3 γ -linolenic 0.15 0.61 1.00 0.25C20:2 0.35 0.37 0.07 0.27C20:3 dihomo-γ -linolenic 0.10 0.10 0.32C20:4 Arachidonic 0.11 0.14 0.46C22:2 0.04 0.11C22:4 0.03 0.09C22:5 0.04 0.09

n-3 Series

C18:3 α-linolenic 8.00 6.32 5.31 1.80 1.03C18:4 0.22C20:3 0.12C20:4 0.07 0.09C20:5 Eicosapentaenoic 0.27 0.02 0.09 0.12C22:5 0.07 0.10 0.19C22:6 Docosahexaenoic 0.30 0.04 0.01 0.25Total PUFA 18.00 16.02 22.77 5.82 15.27Ratio n-6:n-3 1.26 1.17 3.14 2.06 8.09Ratio C18:2 to C18:3 1.26 1.28 2.72 1.55 9.37

Source: Modified from Uniacke and Fox 2011.c, cis; t,trans; i, iso; a, anteiso, PUFA, polyunsaturated fatty acids.

16.7:1. As a species, humans are generally deficient in n-3 fattyacids and have excessive levels of n-6 which is associated withthe pathogenesis of cardiovascular, cancerous, inflammatory andautoimmune diseases (Simopoulos 2002).

Equine milk-fat contains a very low level of conjugatedlinoleic acid (CLA; rumenic acid, Tables 26.9 and 26.10) whichis virtually absent from asinine milk but is high in ruminant milk-fats, being produced in the rumen by abortive biohydrogenationof n-6 octadecadienoic acid (LA) (see Whigham et al. 2000,Bauman and Lock 2006, Collomb et al. 2006). CLA has severaldesirable effects in the diet; some of the positive health effects at-tributed to it include: suppression of carcinogenesis, anti-obesityagent, modulator of the immune system and control of arthero-genesis and diabetes. Small amounts of eicosapentaenoic (EPA)and docosahexaenoic acid (DHA) are present in asinine milkwhereas equine milk has only trace amounts. EPA and DHA areespecially important in infant nutrition but their absence fromequine milk is not considered to be a problem as an infant’s livercan desaturate linoleic and ALA to form EPA and DHA.

Structure of Triglycerides

The distribution of fatty acids in animal TGs is non-random,apparently so that the TGs will be liquid at body tempera-ture. Inter-species comparison of the positional distribution offatty acids has been determined, using fatty acid and stereospe-cific analysis, for 11 species: echidna, koala, Tammar wallaby,guinea pig, dog, cat, Weddell seal, horse, pig, cow and human

(Parodi 1982). Generally, the positional distribution of fatty acidsis similar, except for the echidna, with short chain fatty acidspreferentially esterified at sn-3, saturated fatty acids at sn-1 andunsaturated fatty acids generally at sn-2. In equine milk fat,C10:0 occurs at the sn-3 position, whereas in bovine milk fat,more C10:0 is found at the sn-2 than at the sn-3 position. Inhuman and equine milk, C16:0 is located predominantly at thesn-2 position which is regarded as favourable for assimilationby infants and children, whereas in bovine milk, C16:0 is equallydistributed between the sn-1 and sn-2 positions. In the milk fatof the Weddell seal and horse, C18:1 is esterified preferentiallyat sn-1 but for all other species studied it occurs mainly at sn-3(Parodi 1982). No information is available on the stereospecificdistribution of fatty acids in the TGs of asinine milk.

Equid Milk Fat Globules and MFGM

The fat in milk is emulsified as globules which are surroundedand stabilised by a very complex emulsifying layer, consistingof phospholipids and proteins, called the MFGM. Many of theindigenous enzymes in milk are concentrated in the MFGM.The glycoproteins in the MFGM of human, rhesus monkey,chimpanzee, dog, sheep, goat, cow, grey seal, camel, horse andalpaca have been studied; large intra- and inter-species differ-ences have been found (see Keenan and Mather 2006). Veryhighly glycosylated proteins occur in the MFGM of primates,horse, donkey, camel and dog. Long (0.5–1 µm) filamentousstructures, comprising of mucins (highly glycosylated proteins),

P1: SFK/UKS P2: SFK


514 Part 4: Milk

extend from the surface of the fat globules in equine and humanmilk (Welsch et al. 1988). These filaments dissociate from thesurface into the milk serum on cooling and are lost on heating.For unknown reasons, the filaments on bovine milk fat glob-ules are lost much more easily than those in equine or humanmilk. The filaments facilitate the adherence of fat globules tothe intestinal epithelium and probably improve the digestion offat (Welsch et al. 1988). The mucins prevent bacterial adhe-sion and may protect mammary tissue against tumors (Patton1999). The milk fat globules (MFGs) of asinine milk can be upto approximately 10 µm in diameter, whereas those of equinemilk are generally smaller at 2–3 µm on average; the MFGs ofbovine and human milk are 3–3.5 µm and approximately 4 µm,respectively. Little is known about the proteins of the MFGMof equids but they are known to play a major role in neonataldefense mechanisms in humans (see Mather 2000).

Butyrophilin, acidophilin and XOR have been identified inthe equine MFGM and appear to be similar to the correspond-ing proteins of the human MFGM, as does lactadherin whichshares 74% identity with that of the human lactadherin (Barelloet al. 2008). Both XOR and acidophilin are involved in fat glob-ule secretion with butyrophylin while lactadherin is thought tohave a protective function against rotovirus in the intestinal tract(Barello et al. 2008). Like ovine and buffalo milk, equine milkdoes not cream due to the lack of cryoglobulin.

Rheology Equid Milk Fat

The temperature-dependent melting characteristics of bovinemilk-fat have been studied thoroughly but since equid milk is

not used for the production of butter, the spreadability, rheologyand melting characteristics of these fats have not been studied indetail (see Chandan et al. 1971). Considering the rather unusualfatty acid profile of equid milk fats, they should have interestingmelting and rheological properties.

Stability of Equine Milk Fat

Lipids generally are susceptible to two forms of chemicalspoilage, lipid oxidation (oxidative rancidity) and lipolysis (hy-drolytic rancidity) which are of great commercial significanceto the dairy industry and have been studied in detail (see Foxand McSweeney 2006). No studies on the chemical spoilage ofequid milk-fat have been reported. Considering the high contentof PUFAs in these fats, they are probably quite susceptible to ox-idation. Since equine milk contains a lipase, hydrolytic ranciditywould be expected under certain conditions.

VITAMINSThe overall vitamin content of any milk depends on maternalvitamin status but water-soluble vitamins are more responsive tothe maternal diet than fat-soluble vitamins. Vitamin levels in themilk of some species are shown in Table 26.11. The level of vita-min E is low in asinine milk (∼0.05 mg.L−1) and is reduced fur-ther if the milk is heated. Concentrations of fat-soluble vitaminsare generally similar in equine and bovine milks (Table 26.12).The levels of vitamins A, D3, K and C are significantly higher inequine colostrum than in equine milk, whereas the concentration

Table 26.11. Vitamin Levels (mg.L−1) in the Milk of Some Species

Vitamin Buffalo Goat Sheep Donkey Cow Horse Human

Water-soluble

Thiamine, B1 0.5 0.49 0.48 0.41 0.37 0.3 0.15Riboflavin, B2 1.0 1.5 2.3 0.64 1.8 0.3 0.38Niacin, B3 0.8 3.2 4.5 0.74 0.9 1.4 1.7Pantothenic acid, B5 3.7 3.1 3.5 – 3.5 3 2.7Pyridoxine, B6 0.25 0.27 0.27 – 0.64 0.3 0.14Biotin, B7 0.11 0.039 0.09 – 0.035 – 0.006Folic acid, B9 0.18 0.16Cobalamin, B12 3.0 0.7 0.007a 1.1 0.004 0.003 0.5Ascorbic acid, C – 9.0 4.25b – 21 17.2c 43

Fat-soluble

Vitamin A and β-carotene – 0.5 0.5 – 0.32–0.50 0.12 2.0Cholecalciferol, D3 – – – – 0.003 0.003 0.001α-Tocopherol, E – – – 0.05 0.98–1.28 1.128 6.6Phylloquinone, K – – – – 0.011 0.020 0.002

Source: Modified from Walstra and Jenness 1984, Souci et al. 2000.aRamos et al. 1994.bRecio et al. 2009.cCsapo et al. 1995.

P1: SFK/UKS P2: SFK



Table 26.12. Vitamins (mg.L−1) in Equine Milk

Equine Equine BovineVitamin Colostrum Milk Milk

A 0.88 0.34 0.352D3 0.0054 0.0032 0.0029E 1.342 1.128 1.135K3 0.043 0.029 0.032C 23.80 17.2 15.32

Source: Modified from Csapo et al. 1995.

of vitamin E remains unchanged throughout lactation (Table26.12)(Csapo et al. 1995).

MINERALSMacro-Elements

The levels of ash in equine and asinine milk are similar (Table26.2) and the levels of inorganic elements are close to those

in human milk except for higher concentrations of Ca and P(Table 26.13) (Holt and Jenness 1984). The principal salts inequine milk are phosphates, chlorides, carbonates and citratesof potassium, sodium, calcium and magnesium. However, thereare considerable quantitative inter-species differences in milksalts, (Table 26.13). The concentrations of all macro-elements,except potassium, are higher in equine and asinine milk thanin human milk but all are considerably lower than in bovine,caprine, ovine or porcine milk. The low level of salts in equidmilk reduces renal load, making it suitable in infant nutrition.Pieszka and Kulisa (2005) reported on the low tolerance ofequine species to imbalances in mineral concentrations in milkpost-partum; slight increases in some minerals can cause severedeformation of teeth and bones in horses and affect metabolismand protein synthesis.

The concentration of macro-elements in equine milk is com-parable to that in zebra milk (Equus zebra), whereas in earlylactation, the concentrations of calcium and phosphate are con-siderably higher in domestic horse milk (Equus caballus)(Table26.14) than in zebra milk (Schryver et al. 1986). The concentra-tions of macro-elements in equine milk are strongly influenced

Table 26.13. Total Concentrations of Inorganic Elements (mmol.L−1) and Citrate in the Milk of Eight Different Species

Species Calcium Magnesium Sodium PotassiumPhosphorus(Inorganic) Citrate Chloride

Horse 16.5 1.6 5.7 11.9 6.7 3.1 6.6Cow 29.4 5.1 24.2 34.7 20.9 9.2 30.2Man 7.8 1.1 5.0 16.5 2.5 2.8 6.2Goat 23.1 5.0 20.5 46.6 15.6 5.4 34.2Sheep 56.8 9.0 20.5 31.7 39.7 4.9 17.0Pig 104.1 9.6 14.4 31.4 51.2 8.9 28.7Rat 80.4 8.8 38.3 43.6 93.3 0.06 36.1Rabbit 214.4 19.5 83.7 89.5 54.2 17.4 80.0

Source: Modified from Holt and Jenness 1984.

Table 26.14. Concentration of Inorganic Elements (µg.g−1 Whole Milk) in Early and Late Lactation Milk of VariousEquid Species

Species Total Solids Ash Ca P Mg Na K Cu Zn Fe

Early LactationPrzewalski horse 11.6 0.58 1380 790 104 220 590 0.42 4.1 1.3Hartmann’s zebra 11.3 0.50 1100 800 120 290 590 1.13 2.4 1.1Domestic horse 11.6 0.58 1327 884 102 198 655 0.64 2.7 0.37Domestic pony 9.6 0.48 1036 600 70 189 483 0.26 1.8 –

Late LactationPrzewalski horse 10.3 0.33 804 419 62 137 344 0.23 1.9 1.1Hartmann’s zebra 10.0 0.32 840 550 80 200 422 0.26 1.9 3.6Grant’s zebra 10.7 0.35 690 490 93 277 187 1.0 2.9 2.2Domestic horse 10.2 0.36 811 566 53 140 410 0.25 1.9 0.27Domestic pony 9.4 0.36 857 418 77 127 250 0.37 1.7 –

Source: Modified from Schryver et al. 1986.

P1: SFK/UKS P2: SFK


516 Part 4: Milk

Stage of lactation (wk)181614121086420

Sal

t con

cent

ratio

n (m

g/m

L)

0

200

400

600

800

1000

1200

1400

1600

Figure 26.5. Influence of the stage of lactation on the concentrationof calcium (•), magnesium (◦), phosphate (�), potassium (∇) orsodium (�) in equine milk. (Data from Schryver et al. 1986.)

by the stage of lactation (Fig. 26.5), which shows a progressivedecrease in the concentrations of calcium, magnesium, phos-phorus, sodium and potassium from the end of the first week oflactation. The concentration of calcium in equine milk increasesduring the first week of lactation, before decreasing steadilythereafter (Ullrey et al. 1966). As in the milk from all mammalswhich have been studied in sufficient detail, the concentrationsof calcium and phosphate in equine milk far exceeds the solubil-ity of calcium phosphate in milk. As a result, part of the calciumand phosphate exist in a non-ultrafiltrable, micellar, form, thatis MCP. Holt and Jenness (1984) reported that approximately60%, 20% or 40% of total calcium, magnesium or inorganicphosphorus, respectively, in equine milk are non-ultrafiltrable,compared to approximately 70%, 35% or 45% of total calcium,magnesium and inorganic phosphorus in bovine milk. On thebasis of measurements of the distribution of salts between theultrafiltrable and non-ultrafiltrable phase of milk, Holt and Jen-ness (1984) estimated the concentrations of ionic calcium andmagnesium in equine milk at pH 7.0 are 2.5 and 0.6 mmol L−1,respectively, compared to 2.0 or 0.8 mmol L−1, respectively, inbovine milk at pH 6.7. For human infant nutrition, a Ca:P ratioof approximately 2:1 is considered optimal; for bovine milk, theratio is approximately 1:1, but in equine milk it is about 2:1, andis very close to that in human milk.

Trace Elements

Data on concentrations of trace elements, that is, those elementspresent at concentrations less than 30 mg L−1, in equine milkhave been sporadic. The concentrations of trace elements inequine, bovine and human milk are compared in Table 26.15.Compared to bovine milk, equine milk contains markedly higherlevels of aluminum, copper, iron and titanium but lower lev-els of boron, barium, lithium, molybdenum, manganese, silicon

Table 26.15. Concentrations of Trace Elements(µg.L−1) in Equine, Bovine and Human Milk

Species

Element Equine Bovine Human

Aluminum 123 98 125Boron 97 333 273Barium 76 188 149Copper 155 52 314Iron 224 194 260Lithium 15 24 7Molybdenum 16 22 17Manganese 14 21 7Silicon 161 434 472Strontium 442 417 60Titanium 145 111 25Zinc 1835 3960 2150

Source: Modified from Anderson 1992.

and zinc. Human milk contains more boron, barium, copper,iron, silicon and zinc, but less lithium, manganese, strontiumand titanium than equine milk (Table 26.15). Concentrationsof zinc, iron and copper in equine milk decrease progressivelywith advancing lactation, whereas the concentration of man-ganese increases during the first 5 days of lactation, after whichit decreases progressively (Csapo-Kiss et al. 1995, Ullrey et al.1974).

PHYSICAL PROPERTIES OF EQUID MILKThe physical properties of the milk of some species are comparedin Table 26.16.

Density

The density (kg m−3) of equine colostrum is higher than that ofequine milk (Waelchli et al. 1990, Ullrey et al. 1966, Marianiet al. 2001), due primarily to its considerably higher proteincontent. Values for colostrum can reach up to approximately1080 kg m−3 (Ullrey et al. 1966) and show a significant linearcorrelation with the IgG content of colostrum (Waelchli et al.1990, LeBlanc et al. 1986). Density is highest immediately post-partum and decreases rapidly during the first 12 h (Ullrey et al.1966); considerably smaller decreases in density are observedduring the rest of lactation (Ullrey et al. 1966, Mariani et al.2001). Density values reported for mature equine milk rangefrom approximately 1028 to 1035 kg m−3. The density of wholemature bovine milk normally ranges from 1027 to 1033 kg m−3

(Singh et al. 1997).

Refractive Index

The refractive index for equine colostrum ranges from 1.340to 1.354, whereas that of mature equine milk is approximately

P1: SFK/UKS P2: SFK



Table 26.16. Physical Properties of the Milk of Equid Species, with Comparative Data for Bovine and Human Milk

Property Equine Milk Equine Colostrum Asinine Milk Bovine Milk Human milk

Freezing point (◦C) −0.525 –−0.554

– −0.55 –−0.49

−0.512 –−0.55

–

pH (25◦C) 7.1–7.3 – 7.01–7.35 6.5–6.7 6.8Density (kg.m−3), (20◦C) 1032 1080 1029–1037 1027–1033 1031Refractive index, nD

20 1.3394 1.340–1.354 – 1.344–1.349 –

Viscosity (mPa s) 1.5031 – – 1.6314 –

Zeta potential (mV) −10.3 – – −20.0 –

Colour L∗a∗b∗ 86.52, −2.34, −0.15 – 80.88−2.27−3.53

79.12, −7.46−2.31

–

Source: Modified from Uniacke and Fox 2011.

1.339 (Waelchli et al. 1990). The higher refractive index ofcolostrum than of mature milk is probably related to its highertotal solids content, since the refractive index increases withincreasing mass fraction of each solute. The refractive index ofbovine milk is in the range 1.344–1.349 (Singh et al. 1997).

pH

There is considerable variation among reported values for thepH of equine milk. Mariani et al. (2001) reported that its pH4 days post-partum is approximately 6.6 and increases to ap-proximately 6.9 after 20 days and to approximately 7.1 at 180days post-partum. A value of approximately 7.0 for the pH formature equine milk was reported by Kucukcetin et al. (2003),but Pagliarini et al. (1993) reported an average value of ap-proximately 7.2. The pH of bovine milk is generally between6.5 and 6.7 (Singh et al. 1997) and increases during lactation(Tsioulpas et al. 2007). These differences are presumably re-lated to differences in protein and salt composition of the milks.

Freezing Point

The freezing point of milk is directly related to the concen-trations of water-soluble compounds therein. Fat globules andproteins have a negligible influence on freezing point, with themain effect arising from lactose and minerals. A freezing pointof −0.554◦C (Pagliarini et al. 1993) or −0.548◦C (Neseni et al.1958) has been reported for equine milk, whereas the vast ma-jority of bovine milk samples have a freezing point in the range−0.512 to −0.550◦C (Singh et al. 1997). The lower freezingpoint of equine milk is probably related to its higher lactosecontent.

Viscosity

Equid milks are less viscous than bovine milk due to a lowertotal solids content.

Colour

Equid milk may be expected to be less white than bovine milkdue to its low protein content and large casein micelles but thisis not the case and equid milk is considerably whiter due to theabsence of β-carotene that confers a yellow colour to bovinemilk.

PROCESSING OF EQUID MILKBecause of its unique physico-chemical properties outlined ear-lier, the processing of equine and asinine milk into traditionaldairy products is not possible; cheese is not produced from thesemilks as a firm curd is not formed on renneting. Equid milk willform a weak coagulum under acidic conditions and this is ex-ploited in the production of yoghurt-type products with reputedprobiotic and therapeutic properties. Traditionally, and to date,the only significant product from equine milk is the fermentedproduct, koumiss, the production and properties of which are de-scribed below in section ‘Koumiss’. Interest in koumiss produc-tion has grown recently which may be attributed to the fact that,worldwide, the overall consumption of fermented milk productshas grown faster than the consumption of fresh milk (IDF 2009).

Fermentation is one of the oldest methods for preservingmilk and probably dates back approximately 10,000 years tothe Middle East where the first organised agriculture occurred.Traditional fermented milk products have been developed inde-pendently worldwide and were, and continue to be, especiallyimportant in areas where transportation, pasteurisation and re-frigeration facilities are inadequate. Worldwide, milk from eightspecies of domesticated mammals (cow, buffalo, sheep, goat,camel, horse, reindeer and yak) have been used to produce tradi-tional fermented milk products. There are three categories of fer-mented milks, those resulting from lactic fermentations, yeast-lactic fermentations and mould (Geotrichum candidum)-lacticfermentations. Koumiss and kefir belong to the yeast-lactic fer-mentation group where alcoholic fermentation by yeasts is used

P1: SFK/UKS P2: SFK


518 Part 4: Milk

in combination with a lactic acid fermentation (Tamine and Mar-shall 1984). The conversion of milk into a fermented product hasseveral important advantages; as well as being a means of preser-vation, it also improves taste and digestibility and increases thevariety of food. Current interest in the health benefits of fer-mented milks started with the theory of longevity proposed bythe Russian microbiologist, Elie Metchnikoff (1845–1918); heclaimed that people who consumed fermented milks lived longeras lactic acid bacteria from the fermented product colonised theintestine and inhibited ‘putrefaction’ caused by harmful bacte-ria (a probiotic effect), thereby slowing down the aging process.Nowadays, the principal effects of probiotics are thought to be:improved gastrointestinal transit time of digesta, bowel func-tion and glycemic index; some reports claim that they have ananticarcinogenic effect (see McIntosh 1996).

Further details on the fermentation of milk can be found inmany publications including: Tamine and Marshall (1984); Ko-roleva (1991); Kurmann et al. (1992); Nakazawa and Hosono(1992); Surono and Hosono (2002).

Koumiss

Koumiss (Kumys), fermented equine milk, is widely consumedin Russia, Mongolia and Kazakhstan, primarily for its therapeu-tic value. Russians, in particular, have long advocated the use ofkoumiss for a wide variety of illnesses but the variable micro-biology of these products has made it difficult to confirm anytheoretical basis for the claims (Tamine and Robinson 1999).In Mongolia, koumiss is the national drink (Airag) and a high-alcoholic drink made by distilling koumiss, called Arkhi, is alsoproduced (Kanbe 1992). Per caput consumption of koumiss inMongolia is estimated to be about 50 L/yr.

Koumiss is still manufactured in remote areas of Mongoliaby traditional methods but with increased demand elsewhere itis now produced under more controlled and regulated condi-tions. Traditional koumiss (from fresh raw milk) was preparedby seeding milk with a mixture of bacteria and yeasts using partof the previous day’s product as an inoculum. The milk washeld in a leather sack called a ‘turdusk’ (also called a ‘saba’ or‘burduk’), made from smoked horsehide taken from the thigharea, that is, it has a broad bottom and a narrow, long, sleevewith a capacity of 25–30 L, fermentation takes from 3 to 8hours. In the 1960s, the microbial population was analysed andfound to consist mainly of Lactobacillus delbrueckii subsp. bul-garicus, Lactobacillus casei, Lactococcus lactis subsp. lactis,Kluveromyces fragilis and Saccharomyces unisporus (Tamineand Marshall 1984). The lactic acid bacteria are responsible foracid production and the yeasts for the production of ethanol andcarbon dioxide. During the mixing and maturation stages of pro-duction, more equine milk is usually added to control the levelsof acidity and ethanol. The whole process is poorly controlledand often results in a product with an unpleasant taste due to thepresence of too much yeast or excessive acidification. Turdusks,often containing fermented caprine milk from the previous sea-son, were stored in a cool place over winter and the culture wasreactivated when required by gradually filling the turdusk withequine milk over about 5 days (Tamine and Marshall 1984).

Koumiss contains about 90% moisture, 2–2.5% protein (1.2%casein and 0.9% whey proteins), 4.5–5.5% lactose, 1–1.3% fatand 0.4–0.7% ash, as well as the end-products of microbialfermentation, that is, lactic acid (1.8%), ethanol (0.6–2.5%)and CO2 (0.5–0.9%) and provides 37 to 40 kCal/100 mL). Af-ter production, koumiss contains between 0.6 and 3% ethanoland is effervescent. Koumiss is thought to be more effectivethan raw equine milk in disease treatment due to the additionalpeptides and bactericidal substances produced during microbialmetabolism (Doreau and Martin-Rosset 2002).

In the last decade, technological advances have been made inthe manufacture of koumiss, such as the development of blendsof microorganisms in starter cultures that enhance flavour devel-opment and extend the shelf life up to 14 days. The production ofkoumiss and other fermented milk products is carried out usinga more standardised protocol for manufacture and is of con-siderable interest for increasing the market and consumption ofequine milk products in countries where it has not normally beenconsumed (Di Cagno et al. 2004). As well as pasteurising the rawequine milk, pure cultures of lactobacilli such as Lb. delbrueckiisubsp. bulgaricus and yeasts are used for koumiss manufacture.The use of Saccharomyces lactis is considered best for ethanolproduction (2–5%) and S. cartilaginosus is sometimes used forits antibiotic activity against Mycobacterium tuberculosis (Parket al. 2006). Other microorganisms such as Candida spp., Torulaspp., Lb. acidophilus and Lb. lactis may also be used in koumissproduction (Surono and Hosono 2002). A protocol for the man-ufacture of commercial koumiss is presented in Figure 26.6.The characteristics of a good koumiss are optimal when lac-tic and alcoholic fermentations proceed simultaneously so thatthe products of fermentation occur in definite proportions. Aswell as lactic acid, ethanol and CO2, volatile acids and othercompounds are formed which are important for aroma and tasteand approximately 10% of equine milk proteins are hydrolysed.Products with varying amounts of lactic acid and ethanol are pro-duced and generally 3 categories of koumiss are recognised: mild(0.6–0.8% acidity, 0.7–1.0% alcohol; medium (0.8–1.0% acid-ity, 1.1–1.8% alcohol) and strong (1.0–1.2% acidity, 1.8–2.5%alcohol (Tamine and Marshall 1984).

The presence of a high level of thermostable Lyz in equinemilk may interfere with the activity of some starter microor-ganisms in the production of fermented products and thus causeproblems in the processing of equine milk. Di Cagno et al.(2004) who heated equine milk to 90◦C for 3 minutes to inacti-vate Lyz, produced an acceptable fermented product. In sensorytests, fermented equine milk generally scores low and, in an at-tempt to enhance the rheological and sensory properties of fer-mented products made from equine milk, Di Cagno et al. (2004)fortified equine milk with bovine Na caseinate (1.5 g.100−1g),pectin (0.25 g.100−1g) and threonine (0.08 g.100−1g). The re-sultant product had good microbiological, rheological and sen-sory characteristics after 45 days at 4◦C. Fermented unmodifiedequine milk had an unacceptable viscosity and scored very lowin comparison to fortified products for appearance, consistencyand taste.

Research has turned also to producing koumiss-like productsfrom bovine milk, which must be modified to make it suitable for

P1: SFK/UKS P2: SFK



Bovine milk, skim andpasteurise at 70oC

x 30 min

Cool part to 26oC,inoculate with

yeast culture (e.g. Torula spp.)

Cool part to 40oC, inoculate

with lactic culture (e.g. L. delbrueckiisubsp. bulgaricus)

Plug with cotton wool, mix and

Plug and mix, hold for 6–7 h at 35–37oC

hold for 15–18 h at 28–30oC

Mother culture

Mix and pour into double-walled starter vat with agitator through cover. Add

fresh mares’ milk continuously at 30oC, to acidity of ~0.7% lactic acid. Agitate for 15 min & ripen at 26–28oC.

Allow to ferment for 3–4 days to ~1.4% lactic acid

Bulk starter

Fresh mares’ milk at 26–28oC indouble-walled vat. Inoculate at a rate

of 30% bulk starter to 0.5% lactic acid, incubate and mix for ~60 min to 0.6%

lactic acid.Bottle and cap, hold for 1–2 h at 18–20oC.

Store for 12–24 h at 4–6oCKoumiss

Figure 26.6. Schematic for the production of koumiss. (Based onBerlin 1962.)

koumiss production. Methods have been developed, with vary-ing degrees of success, where a single constituent of bovine milkhas been altered to resemble that of equine milk, for example thecarbohydrate content has been increased or the protein contentreduced but, until recently both had not been altered simulta-neously. Koumiss of reasonable quality has been produced suc-cessfully from whole or skimmed bovine milk containing addedsucrose using a mixture of Lb. acidophilus, Lb. delbrueckii ssp.

bulgaricus and Kluyveromyces marxianus var. marxianus or var.lactis as starter culture (Kucukcetin et al. 2003). Koumiss hasalso been made from diluted bovine milk supplemented withlactose and, more successfully, from bovine milk mixed withconcentrated whey using a starter culture of Kluyveromyces lac-tis (AT CC 56498), Lb. delbrueckii subsp. bulgaricus and Lb.acidophilus. Starter cultures for koumiss manufacture frombovine milk may also include Saccharomyces lactis (highantimicrobial activity against Mycobacterium tuberculosis) inorder to retain the ‘anti-tuberculosis image’ of equine milk(Kucukcetin et al. 2003). More recently, bovine milk has beenmodified to approximate the composition of mares’ milk us-ing a series of membrane filtration steps and a starter culture(Kluyveromyces lactis, Lb. delbrueckii subsp. bulgaricus andLb. acidophilus) that ensures consistent fermentation; the re-sulting product was found to be very similar to koumiss withrespect to pH, titratable acidity, ethanol content, proteolytic ac-tivity, apparent viscosity and microbial composition, both whenfresh or stored (15 days at 4◦C) (Kucukcetin et al. 2003).

The physico-chemical and microbiological properties of asi-nine milk, similar to equine milk, such as low microbiologicalload and high Lyz make it a good substrate for the production offermented products with probiotic Lactobacillus strains. Cop-pola et al. (2002) incubated asinine milk with the probiotic Lb.rhamnosus (AT 194, GTI/1, GT 1/3) and found that the strain isunaffected by the high Lyz activity in the milk and remained vi-able after 15 days at 4◦C and pH 3.7–3.8. Lb. rhamnosus inhibitsthe growth of most harmful bacteria in the intestine and acts asa natural preservative in yoghurt-type products, considerablyextending shelf life. Chiavari et al. (2005) produced fermentedbeverages from asinine milk using a mixed culture of Lb. rham-nosus (AT 194, CLT 2.2) and Lb. casei (LC 88) and in all casesfound a high level of viable bacteria after 30 days storage. Somesensory differences were recorded for the fermented drinks andthose made with the Lb. casei strain developed a more accept-able and balanced aroma than the boiled vegetable/acidic tasteand aroma of the products made with Lb. rhamnosus.

Other Products from Equid Milk

As sales of equine milk have increased considerably in recentyears, research is now focused on the development of new fer-mented products or new methods for extending the shelf life ofexisting products, while maintaining some of the unique compo-nents of equine milk. The ability of milk to withstand relativelyhigh processing temperatures is very important from a techno-logical point of view. The whey proteins in equine milk are muchmore thermostable than those of bovine milk. Heat treatment at80◦C × 80 s causes only a 10–15% decrease in non-caseinnitrogen, with a marked decrease evident only when the tem-perature is increased above 100◦C (Bonomi et al. 1994). Lf andequine BSA are the most heat-sensitive but are not completelydenatured until the temperature reaches 130◦C x 10 min. β-Lgand α-La are almost completely denatured at temperatures over130◦C and Lyz at temperatures greater than 110◦C (68% residualLyz activity after heating at 82◦C for 15 minutes (Jauregui-Adell1975).

P1: SFK/UKS P2: SFK


520 Part 4: Milk

NUTRITIONAL AND BIOMEDICALPROPERTIES OF EQUID MILKIt is claimed that fresh and fermented equid milk relievemetabolic and intestinal problems while having a gut-cleansingeffect coupled with ‘repair’ of the intestinal micro-flora. It isclaimed to give relief from stomach ulcers, high blood pressure,high cholesterol and liver problems and is also recommendedas an aid in the treatment of cancer patients. The recommendedamount of equine milk is 250 mL/d. The most significant useof equid milk is as a substitute for bovine milk for patientswith cows’ milk protein allergy (CMPA); limited research hasshown that both equine and asinine milks are generally toleratedwell by CMPA patients. The use of equine milk in the produc-tion of cosmetics is relatively new and includes soaps, creamsand moisturisers (Doreau and Martin-Rosset 2002). At present,equine milk is available in several forms: frozen milk, frozenyoghurt-type drink, lyopholised powder, shampoos and variouscosmetic and medicinal creams. Lyopholised equine colostrumis available and used in the high-value horse industry to feed or-phaned foals. Marconi and Panfili (1998) suggested that researchis required to identify optimum drying and storage conditions forpowdered equine milk for retention of some of the unique char-acteristics of raw milk, including high levels of whey proteins,PUFAs, lysine, tocopherols and vitamin C which are partially de-stroyed in the preparation of commercial powdered equine milk.Asinine milk is used to make ice cream and other desserts andalso a fermented milk product. To improve the nutritive value ofasinine milk and increase its overall energy content for humannutrition, it is frequently supplemented with approximately 4%medium-chain TGs (Salimei 2011).

Cow Milk Protein Allergy

Equine and asinine milk, with a composition close to that ofhuman milk, may be good nutritional sources for the neonatewhen breast milk is not available. Bovine milk or bovine milkproducts are used traditionally as substitutes for human milk ininfant nutrition but bovine milk is considerably different fromhuman milk in terms of its macro- and micro-nutrients and theabsorption rates of vitamins and minerals from the two milksare different, which can be problematic for infants. CMPA isan IgE-mediated type I allergy, which may be life-threatening,and is defined as a set of immunologically mediated adversereactions which occur following the ingestion of milk, affectingfrom 2 to 6% of children in their first year of life. About 50%of affected children recover after the age of one and 80–90% ofthose affected recover by 5 years of age (Caffarelli et al. 2010).The high frequency of CMPA in infants and children is thoughtto be due to an incomplete gut mucosal barrier, increased gutpermeability to large molecules and immature local and systemicresponses which are aided by breast-milk which facilitates gutmaturation and provides passive protection against bacteria andantigens (Hill 1994).

The difference between bovine milk protein allergy and lac-tose intolerance is of particular interest and it is an area whichcauses much confusion. CMPA is a food allergy, that is an ad-

verse immune reaction to a food protein that is normally harm-less to the non-allergic individual. Lactose intolerance is a non-allergic food hypersensitivity due to a deficiency of the enzymeβ-galactosidase (lactase), required to hydrolyse lactose. Lactasedeficiency manifests as abdominal symptoms and chronic diar-rhoea after ingestion of milk (see Bindslev-Jensen 1998, Vesaet al. 2000). Lactose intolerance is not a disease or malady;70% of the world’s population is lactose-intolerant. Adverse ef-fects of lactose intolerance occur at a much higher level of milkconsumption than that which causes milk allergy.

CMPA is important because bovine milk is the first foreignantigen ingested in large quantities in early infancy. Reviews onCMPA include: Hill (1994); Hill and Hosking (1996); Taylor(1986); Høst (1988, 1991); Bindslev-Jensen (1998); Wal (2002,2004); El-Agamy (2007); Apps and Beattie (2009). Becauseβ-Lg is absent from human milk, it has commonly been con-sidered to be the most important cows’ milk allergen (Goldmanet al. 1963, Ghosh et al. 1989) although other whey proteins(Jarvinen et al. 2001) and caseins (Savilahti and Kuitunen 1992,Restani et al. 1995) have also been implicated in allergic reac-tions. In children, β-Lg is the major allergen, whereas caseinappears to be the most allergenic for adults. The resistance ofβ-Lg to proteolysis allows the protein to remain intact throughthe gastrointestinal tract with the possibility of being absorbedacross the gut mucosa. Ingested β-Lg has been detected in hu-man milk and could be responsible for colic in breast-fed infantsand the sensitisation of infants, predisposing them to allergies(Jakobsson and Lindberg 1978, Kilshaw and Cant 1984, Stuartet al. 1984, Jakobsson et al. 1985, Axelsson et al. 1986).

The choice of substitute for cows’ milk in cases of CMPAdepends on two major factors, that is, nutritional adequacy andallergenicity; cost and taste must also be taken into account.Many soy or hydrolysate (casein-based, and more recently, wheyprotein-based) formulae are available for treatment of CMPA butthey can themselves induce allergic reactions. Heat treatment ofmilk may destroy heat-labile proteins, especially BSA and Igs,and change the antigenic properties of other whey proteins, suchas β-Lg and α-la, although caseins need severe heat treatment(121◦C × 20 minutes) to reduce sensitising capacity (Hill 1994).Enzymatic treatment of milk proteins may result in productswith an unacceptable taste due to bitterness arising from theproduction of peptides and amino acids and such peptides may,in fact, be allergenic (Schmidt et al. 1995, Selo et al. 1999,El-Agamy 2007).

Many clinical studies have been carried out on the use of themilk of different species in infant nutrition, for example goat,sheep (Restani et al. 2002), camel (El-Agamy 2007), buffalo(El-Agamy 2007), horse and donkey (Iacono et al. 1992,Carroccio et al. 2000, El-Agamy et al. 1997, Businco et al. 2000,Monti et al. 2007). Results on the benefits of such milks are con-flicting and infants with CMPA may suffer allergic reactions tobuffalo, goat, sheep, donkey or mare milk proteins due to positiveimmunological cross-reaction with their counterparts in cows’milk (El-Agamy 2007). Lara-Villoslada et al. (2005) found thatthe balance between casein and whey proteins may be impor-tant in determining the allergenicity of bovine milk proteins inhumans and that modification of this balance may reduce the

P1: SFK/UKS P2: SFK



allergenicity of bovine milk; a readjustment of the casein:wheyproteins ratio to 40:60 was found to make the bovine milk sig-nificantly less allergenic. Presumably, equine and asinine milk,with a ratio of casein:whey proteins close to that in human milk,are potentially good substitutes for human milk. It is notewor-thy that the study by Lara-Villoslada et al. (2005) was carriedout using mice as subjects and the increased level of β-Lg (amajor bovine milk allergen) in bovine milk adjusted to 40:60,casein:whey was not considered.

Cross-Reactivity of Milk Proteins

Cross-reaction occurs when two food proteins have similaramino acid sequences or when the three-dimensional confor-mation makes two molecules similar in their capacity to bindspecific antibodies (Restani et al. 2002). Cross-reactivity of pro-teins from different species generally reflects the phylogeneticrelations between animal species, for example homologous pro-teins from vertebrates often cross-react. A comprehensive studyon the subject by Jenkins et al. (2007) highlights some interesting

points, especially concerning the potential allergenicity of ca-seins from different species. The authors set out to determinehow closely a foreign protein has to resemble a human homo-logue before it actually loses its allergenic affect. A high degreeof similarity to human homologues would, presumably, implythat a foreign animal food protein would be much less likely thana protein with little or no similarity to its human homologue to beallergenic in human subjects. In addition, the study of potentialanimal allergens must take into account the ability of the humanimmune system to discriminate between its own proteins, that isan autoimmune response, and those from another species whichhave a high similarity, that is how closely a foreign protein hasto resemble a human homologue before it loses its ability toact as an allergen? (Spitzauer 1999). Table 26.17 gives the per-centage homology of αS1-, α-S2 and β-caseins from differentspecies to bovine and human homologues. Known allergens areless than 53% identical to human sequences. Natale et al. (2004)found that 90% of a group of infants with CMPA had serum IgEagainst bovine αS2-casein, 55% against bovine αS1-casein andonly 15% against bovine β-casein, which is closest in amino acid

Table 26.17. Homology (percentage) Between Milk Proteins from DifferentSpecies and Their Human Homologues

Percent Identity to Closest

Casein Accession Code Bovine Homolog Human Homolog

α-S1-Casein

Cow P02662 100 29Goat Q8M1H4 88 29Sheep P04653 88 28Horse Q8SPR1 39 44Human P47710 29 100Rat PO2661 22 27Camel O97943 41 36Rabbit PO9115 37 37

α-S2-Casein

Cow P02663 100 16Goat P33049 88 17Sheep P04654 89 17Camel O97944 56 11Rabbit P50418 36 16

β-Casein

Cow P02666 100 53Goat Q712N8 91 54Sheep P11839 91 54Horse Q9GKK3 56 58Human P05814 53 100Camel Q9TVD0 66 58Rabbit PO9116 52 55

Source: Modified from Jenkins et al. 2007.

P1: SFK/UKS P2: SFK


522 Part 4: Milk

composition to human β-casein. Caprine and ovine milk pro-teins are more closely related to each other than either is tobovine milk proteins, thus explaining why an individual aller-gic to goats’ milk cheese may exhibit high IgE cross-reactivitywith sheep’s milk proteins but could tolerate cow’s milk and itsproducts.

Allergy to equine milk appears to be rare and, to date, only twodocumented cases have been reported. Fanta and Ebner (1998)reported the case of an individual who experienced sensitisa-tion to horse dander allergen and subsequently produced IgEantibodies on ingestion of equine milk which was prescribed to‘strengthen’ her immune system. Gall et al. (1996) demonstratedthe existence of an IgE-mediated equine milk allergy in one pa-tient, caused by low MW heat-labile proteins, most likely α-Laand β-Lg, without cross-reaction to the corresponding wheyproteins from bovine milk. Presumably, the previous cases arenot isolated incidents and as the consumption of equine milkand its products increases, it is likely that further cases will bereported.

Bevilacqua et al. (2001) tested the capacity of goats’ milkwith low or high αs1-casein content to induce milk protein sen-sitisation in guinea pigs and found significantly less sensitisationby milk with low αs1-casein. This may represent an importantattribute of the low αs1-casein content of equine milk for use inhuman allergology. The absence of αs2-casein (and lack of αs1

casein in one donkey) and β-lg II in donkey milk reported byCriscione et al. (2009) could be potentially interesting for futureresearch on the allergenicity of asinine milk; αs1-casein and β-lgare scarce or absent in human milk and are considered to be themost significant proteins causing allergic reactions in childrenand adults.

SUMMARYThe characteristics of equine and asinine milk of interest in hu-man nutrition include an exceptionally high concentration ofpolyunsaturated fatty acids, low cholesterol content, high lac-tose and low protein levels (Solaroli et al. 1993, Salimei et al.2004), as well as high levels of vitamins A, B and C. The lowfat and unique fatty acid profile of both equine and asinine milkresults in low atherogenic and thrombogenic indices. Researchhas shown that human health is considerably improved when di-etary fat intake is reduced and, more importantly, when the ratioof saturated to unsaturated fatty acids is reduced. The high lac-tose content of equid milk gives good palatability and improvesintestinal absorption of calcium, which is important for bonemineralisation in children. The renal load of equine milk, basedon levels of protein and inorganic substances, is equal to that ofhuman milk, a further indication of its suitability as an infantfood. Equine and asinine milk can be used for their prebioticand probiotic activity and as alternatives for infants and childrenwith CMPA and multiple food intolerances (Iacono et al. 1992,Carroccio et al. 2000).

The invigorating effect of equine milk may be, at least par-tially, due to its immunostimulating ability. Lyz, Lf and n-3fatty acids have long been associated with the regulation of

phagocytosis of human neutrophils in vitro (Ellinger et al. 2002).The concentration of these compounds is exceptionally high inequine milk and the consumption of frozen equine milk signifi-cantly inhibits chemotaxis and respiratory burst, two importantphases of the phagocytic process (Ellinger et al. 2002). Thisresult suggests a potential anti-inflammatory effect by equinemilk.

To be successful as a substitute for human milk in infantnutrition, equine milk must be capable of performing many bi-ological functions associated with human milk. The presence ofhigh concentrations of Lf, Lyz, n-3 and n-6 fatty acids in equinemilk are good indicators of its potential role. However, the lackof research must be addressed to develop the potential of equinemilk in the health and nutritional markets. Studies are requiredto bring the health claims for equine milk out of the realms of re-gional folklore. It seems reasonable to suggest that equine milkcould be marketed as a dietary aid where the immune systemis already depleted, that is, as a type of ‘immuno-boost’. Morethan 30% of customers who purchase equine milk in the Nether-lands are patients undergoing chemotherapy, who find equinemilk helpful in counteracting the effects of the treatment. Thecomposition of equine milk suggests a product with interest-ing nutritional characteristics with potential use in dietetics andtherapeutics, especially in diets for the elderly, convalescent andnewborns.

Future research should also include comprehensive character-isation of the proteins of different breeds of horse and donkey,with the possibility of selection of animals for specific proteinswhich could, in turn, optimise the nutritional and technologi-cal properties of the milks; the genetic control of αs-, β- andκ-caseins as well as β-Lg in bovine and caprine milk has beenresearched, albeit in a somewhat limited manner, as has the ef-fects of these proteins on the technological properties of bothtypes of milk, for example colloidal stability, coagulation andcurd strength. These characteristics can determine the physicaland chemical behaviour of milk in the infant gastrointestinaltract with the result that digestion and availability of nutrientsto the young may be affected (Cuthbertson 1999). Furthermore,genetic selection of certain breeds of horse and donkey may im-prove milk yield and lactation pattern and make the productionof milk more cost effective.

REFERENCES

Addeo F et al. 1984. Susceptibility of buffalo and cow κ-caseins tochymosin action. Milchwissenschaft 39: 202–205.

Aganga AA et al. 2000. Feeding donkeys. Livest Res Rural Dev 12:1–5.

Agostini C et al. 2000. Free amino acid content in standard infantformulas: comparison with human milk. J Am Coll Nutr 19:434–438.

Anderson RR. 1992. Comparison of trace elements in milk of fourspecies. J Dairy Sci 75: 3050–3055.

Apps JR, Beattie RM. 2009. Cows milk allergy in children. BMJ339: 343–345.

Atkinson SA et al. 1989. Non-protein nitrogen components in hu-man milk: biochemistry and potential roles. In: SA Atkinson,

P1: SFK/UKS P2: SFK



B Lonnerdal (eds.) Protein and Non-protein-nitrogen in HumanMilk, Chapter 12. CRC Press, Boca Raton, FL, pp. 117–136.

Atkinson SA, Lonnerdal B. 1995. Nitrogenous components of milk.Non protein nitrogen fractions of human milk. In: RG Jensen (ed.)Handbook of Milk Composition, Chapter 5B. Academic Press,San Diego, CA, pp. 369–387.

Axelsson I et al. 1986. Bovine β-lactoglobulin in human milk. ActaPaediatrica Scandinavica 75: 702–707.

Baldi A et al. 2005. Biological effects of milk proteins and theirpeptides with emphasis on those related to the gastrointestinalecosystem. J Dairy Res 72: 66–72.

Barello C et al. 2008. Analysis of major proteins and fat fractionsassociated with mare’s milk fat globules. Mol Nutr Food Res 58:1448–1456.

Bauman DE, Lock AL. 2006. Conjugated linoleic acid: Biosynthesisand nutritional significance. In: PF Fox, PLH McSweeney (eds.)Advanced Dairy Chemistry, vol 2: Lipids, 3rd edn, Springer, NewYork, pp. 93–136.

Belizy S et al. 2001. Changes in antioxidative properties of lactofer-rin from women’s milk during deamidation. Biochem (Moscow)66: 576–580.

Bell K et al. 1981a. Equine whey proteins. Comp Biochem PhysiolB Biochem Mol Biol 68: 225–236.

Bell K et al. 1981b. Bovine alpha-lactalbumin C and alpha S1-,beta- and kappa-caseins of Bali (Banteng) cattle, Bos (Bibos)javanicus. Aust J Biol Sci 34: 149–159.

Bellamy W et al. 1992. Identification of the bactericidal domain oflactoferrin. Biochimica et Biophysica Acta 1121: 130–136.

Benirschke K et al. 1964. Chromosomes studies of a donkey-Grevyzebra hybrid. Chromosoma 15: 1–13.

Berlin PJ. 1962. Kumiss. In: Annual Bulletin, IV. International DairyFederation, Brussels, pp. 4–16.

Bernhart FW. 1961. Correlations between growth-rate of the suck-ling of various species and the percentage of total calories fromprotein in the milk. Nature 191: 358–360.

Bevilacqua C et al. 2001. Goats’ milk of defective αs1-caseingenotype decreases intestinal and systemic sensitization toβ-lactoglobulin in guinea pigs. J Dairy Res 68: 217–227.

Bindslev-Jensen C. 1998. Food allergy. BMJ 316: 1299–1302.Bonomi F et al. 1994. Thermal sensitivity of mares’ milk proteins.

J Dairy Res 61: 419–422.Boulot S. 1987. L’ingestion chez la jument. Etude de quelques

facteurs de variation au tours du cycle gestation-lactation. Impli-cations, nutritionelles et metaboliques. These Doct. Ing., ENSARennes, University of Rennes I, France.

Brew K. 2003. α-Lactalbumin. In: PF Fox, PLH McSweeney(eds.) Advanced Dairy Chemistry, vol 1: Proteins, 3rd edn,Kluwer Academic/Plenum Publishers, New York, pp. 387–419.

Brock JH. 1997. Lactoferrin structure – function relationships.In: TW Hutchens, B Lonnerdal (eds.) Lactoferrin: Interac-tions and Biological Functions, Chapter 1. Humana Press, NJ,pp. 3–38.

Bruns E. 1999. Commission on horse production. Livest Prod Sci60: 197–198.

Buchheim W et al. 1989. Verrgleichende Untersuchungen zur Struk-tur und Grobe von Casein Micellen in der MilchverschiedenerSpecies. Kieler Milchwirtschaffliche Forschungsberichte 41:253–265.

Businco L et al. 2000. Allergenicity of mare’s milk in children withcow’s milk allergy. J Allergy Clin Immunol 105: 1031–1034.

Caffarelli C et al. 2010. Cow’s milk protein allergy in children: apractical guide. Ital J Pediatr 36: 1–7.

Carbone L et al. 2006. Evolutionary movement of centromeres inhorse, donkey, and zebra. Genomics 87: 777–782.

Carlson SE. 2001. Docosahexaenoic acid and arachidonic acid ininfant development. Semin Neonatol 6: 437–449.

Carratu B et al. 2003. Nitrogeneous components of human milk:non-protein nitrogen, true protein and free amino acids. FoodChem 81: 357–362.

Carroccio A et al. 2000. Intolerence to hydrolysed cow’s milk pro-teins in infants: clinical characteristics and dietary treatment. ClinExp Allergy 30: 1597–1603.

Casey CE. 1989. The nutritive and metabolic advantages of homol-ogous milk. Proc Nutr Soc 48: 271–281.

Cavaletto M et al. 1994. Human αs1-casein like protein: purificationand N-terminal sequence determination. Biol Chem Hoppe Seyler375: 149–151.

Chandan RC et al. 1964. Lysozyme content of human milk. Nature224: 76–77.

Chandan RC et al. 1971 Physicochemical analyses of bovine milkfat globule membrane. I. Differential thermal analysis. J DairySci 54: 1744–1751.

Chen Y et al. 2010. Identification of angio 1-converting enzymeinhibitory peptides from koumiss, a traditional fermented mare’smilk. J Dairy Sci 93: 884–892.

Chianese L et al. 2010. Proteomic characterization of donkey milk“caseome”. J Chromatogr A1 217: 4834–4940.

Chiavari C et al. 2005. Use of donkey’s milk for a fermented bever-age with lactobacilli. Lait 85: 481–490.

Chilliard Y, Doreau M. 1985. Characterization of lipase in maremilk. J Dairy Sci 68: 37–39.

Chowanadisai W et al. 2005. Detection of a single nucleotide poly-morphism in the human α-lactalbumin gene: implications forhuman milk proteins. J Nutr Biochem 16: 272–278.

Chowdhury FA et al. 2004. Protein dissection experiments revealkey differences in the equilibrium folding of alpha-lactalbuminand the calcium binding lysozymes. Biochem 43: 9961–9967.

Civardi G et al. 2002. Capillary electrophoresis (CE) applied toanalysis of mare’s milk. Milchwissenschaft 57: 515–517.

Civardi G et al. 2007. Mare’s milk: monitoring the effect of ther-mal treatments on whey proteins stability by SDS capillary elec-trophoresis (CE-SDS). Milchwissenschaft 62: 32–35.

Clark S, Sherbon JW. 2000. Alphas1-casein, milk composition andcoagulation properties of goat milk. Small Ruminant Res 38:123–134.

Clutton-Brock J. 1992. Horse Power: A History of the Horse andDonkey in Human Societies. British Museum, London.

Collomb M et al. 2006. Conjugated linoleic acids in milk fat:Variation and physiological effects. Intern Dairy J 16: 1347–1361.

Conneely O. 2001. Antiinflammatory activities of lactoferrin. J AmColl Nutr 20: 389S–395S.

Coppola R et al. 2002. Behaviour of L. rhamnosus strain in ass’smilk. Ann Microbiol 52: 55–60.

Creamer L, Sawyer W. 2011. Beta-lactoglobulin. In: J Fuquay et al.(eds.) Encyclopedia of Dairy Sciences, 2nd edn. Elsevier, Oxford,in press.

P1: SFK/UKS P2: SFK


524 Part 4: Milk

Criscione A et al. 2009. Donkey’s milk protein fraction investigatedby electrophoretic methods and mass spectrometric analysis. In-tern Dairy J 19: 190–197.

Csapo J et al. 1995. Composition of mare’s colostrum and milk. Fatcontent, fatty acid composition and vitamin content. Intern DairyJ 5: 393–402.

Csapo-Kiss Zs et al. 1995. Composition of mare’s colostrum andmilk. Protein content, amino acid composition and contents ofmacro- and micro-elements. Intern Dairy J 5: 403–415.

Cunsolo V et al. 2007. Characterization of the protein profile ofdonkey’s whey fraction. Mass Spectr 42: 1162–1174.

Curadi MC et al. 2001. Use of mare milk in pediatric allergol-ogy. Proceedings of the XIV Congress Associazione Scien-tifica Produzioni Animali. Firenze, Italy. Wageningen Academic,Wageningen, The Netherlands.

Cuthbertson WFJ. 1999. Evolution of infant nutrition. Brit J Nutr81: 359–371.

Dalgleish DG. 1998. Casein micelles as colloids. Surface structuresand stabilities. J Dairy Sci 81: 3013–3018.

Davis TA et al. 1994. Amino acid composition of human milk is notunique. J Nutr 124: 1126–1132.

De Kruif CG, Holt C. 2003. Casein micelle structure, func-tions and interactions. In: PF Fox, PLH McSweeney (eds.)Advanced Dairy Chemistry, vol 1: Proteins, 3rd edn, Chap-ter 5. Kluwer Academic/Plenum Publishers, New York, pp. 233–276.

De Kruif CG, Zhulina EB. 1996. κ-Casein as a polyelectrolyte brushon the surface of casein micelles. Colloids Surf A PhysicochemEng Asp 117: 151–159.

Dev BC et al. 1994. κ-Casein and β-casein in human milk micelles:structural studies. Arch Biochem 314: 329–336.

Di Cagno R et al. 2004. Uses of mares’ milk in the manufacture offermented milk. Intern Dairy J 14: 767–775.

Dils RR. 1986. Comparative aspects of milk fat synthesis. J DairySci 69: 904–910.

Doi H et al. 1979. Heterogeneity of reduced bovine κ-casein.J Dairy Sci 62: 195–203.

Donnet-Hughes A et al. 2000. Bioactive molecules in milk andtheir role in health and disease: The role of transforming growthfactor-beta. Immunol Cell Biol 78: 74–79.

Donovan S. 2009. Human milk oligosaccharides – the plot thickens.Brit J Nutr 101: 1267–1269.

Doreau M et al. 1986. Relationship between nutrient intake, growthand body composition of the nursing foal. Reprod Nutr Dev 26:683–690.

Doreau M, Boulot S. 1989. Recent knowledge on mare milk pro-duction: A review. Livest Prod Sci 22: 213–235.

Doreau M et al. 1990. Yield and composition of milk from lactatingmares: effect of lactation stage and individual differences. J DairyRes 57: 449–454.

Doreau M. 1994. Le lait de jument et sa production: particulariteset facteurs de variation. Lait 74: 401–418.

Doreau M, Martin-Rosset W. 2002. Dairy animals: “Horse”. In: HRoginski et al.(eds.) Encyclopedia of Dairy Sciences. AcademicPress, London, pp. 630–637.

Duggan VE et al. 2008. Amyloid A in equine colostrum and earlymilk. Vet Immunol Immunopathol 121: 150–155.

Dziuba J, Minkiewicz P. 1996. Influence of glycosylationon micelle-stabilizing ability and biological properties of

C-terminal fragments of cow’s κ-casein. Intern Dairy J 6: 1017–1044.

Egito AS et al. 2001. Susceptibility of equine κ- and β-caseins tohydrolysis by chymosin. Intern Dairy J 11: 885–893.

Egito AS et al. 2002. Separation and characterization of mares’ milkαs1-, β-, κ-caseins, γ -casein-like, and proteose peptone compo-nent 5-like peptides. J Dairy Sci 85: 697–706.

Egito AS et al. 2003. Action of plasmin on equine β-casein. InternDairy J 13: 813–820.

El-Agamy EI et al. 1997. A comparative study of milk proteins fromdifferent species. II. Electrophoretic patterns, molecular charac-terization, amino acid composition and immunological relation-ships. In: Third Alexandria Conference on Food Science andTechnology. Alexandria, Egypt, pp. 51–62.

El-Agamy EI. 2007. The challenge of cow milk protein allergy.Small Ruminant Res 68: 64–72.

Ellinger S et al. 2002. The effect of mare’s milk consumptionon functional elements of phagocytosis of human neutrophilsgranulocytes from healthy volunteers. Food Agric Immunol 14:191–200.

Ellison RT, Giehl TJ. 1991. Killing of Gram-negative bacteria bylactoferrin and lysozyme. JClin Invest 88: 1080–1091.

El-Zahar K et al. 2005. Peptic hydrolysis of ovine β-lactoglobulinand α-lactalbumin. Exceptional susceptibility of native ovineβ-lactoglobulin to pepsinolysis. Intern Dairy J 15: 17–27.

Enbergs H et al. 1999. Eiweiss-, Glucose- und Lipidgehalt in derMilch von Zuchtstuten im Verlauf der Laktation. Zuchtungskunde71: 245–266.

ExPASy ProtParam Tool, Swiss Institute of Bioinfomatics. 2009.Available at http://www.expasy.org/tools/pi_tool.html. Accessedon October 16, 2009.

Fanta C, Ebner C. 1998. Allergy to mare’s milk. Allergy 53:539–540.

Farrell HM et al. 2006. Casein micelle structure: what can be learnedfrom milk synthesis and structural biology? Curr Opin ColloidInterface Sci 11: 135–147.

Fiat AM et al. 1980. Localisation and importance of the sugar partof human casein. Eur J Biochem 111: 333–339.

Fitzgerald RJ, Meisel H. 2003. Milk protein hydrolysates andbioactive peptides. In: PF Fox, PLH McSweeney (eds.). Ad-vanced Dairy Chemistry, vol 1: Proteins, 3rd edn. Kluwer Aca-demic/Plenum Publishers, New York, pp. 675–698.

Fitzgerald RJ. Murray BA. 2006. Bioactive peptides and lactic fer-mentations. Int J Dairy Technol 59: 118–125.

Formaggioni P et al. 2003. Casein number variability of mare milkfrom Haflinger and Italian Saddle horse breeds. Annali della Fa-colta di Medicina Veterinaria, Universita di Parma 13: 175–179.

Fox PF (ed.). 1985. Developments in Dairy chemistry, vol 3: Lactoseand Minor Constituents. Elsevier Applied Science, London.

Fox PF (ed.). 1997. Advanced Dairy Chemistry, vol 3: Lactose,Water, Salts and Vitamins, 2nd edn. Chapman and Hall, London.

Fox PF, McSweeney PLH. 1998. Dairy Chemistry and Biochem-istry. Blackie Academic and Professional, New York.

Fox PF, Kelly A. 2006. Indigenous enzymes in milk: Overview andhistorical aspects-Part 1. Intern Dairy J 16: 500–516.

Fox PF, McSweeney PLH (eds.). (2006). Advanced Dairy Chem-istry, vol 3: Lipids, 3rd edn. Springer, New York.

Fox PF, Brodkorb A. 2008. The casein micelle: historical aspects,current concepts and significance. Intern Dairy J 18: 677–684.

P1: SFK/UKS P2: SFK



Gall H et al. 1996. Allergy to the heat-labile proteins α-lactalbuminand β-lactoglobulin in mare’s milk. J Allergy Clin Immunol 97:1304–1307.

Ghosh J et al. 1989. Hypersensitivity of human subjects to bovinemilk proteins: a review. Indian J Dairy Sci 42: 744–749.

Girardet J-M et al. 2004. Multiple forms of equine α-lactalbumin:evidence for N-glycosylated and deamidated forms. Inter DairyJ 14: 207–217.

Girardet J-M et al. 2006. Determination of the phosphorylation leveland deamidation susceptibility of equine β-casein. Proteomics 6:3707–3717.

Giuffrida MG et al. 1992. The amino acid sequence of two iso-forms of alpha-lactalbumin from donkey (Equus asinus) milk isidentical. Biol Chem Hoppe Seyler 373: 931–935.

Godovac-Zimmermann J et al. 1985. The amino acid sequence ofβ-lactoglobulin II from horse colostrum (Equus caballus, Peris-sodactyla): β-lactoglobulins are retinol-binding proteins. BiolChem Hoppe Seyler 366: 601–608.

Godovac-Zimmermann J et al. 1987. Identification and the primarystructure of equine α-lactalbumin B and C (Equus caballus, Peris-sodactyla). Biol Chem Hoppe Seyler 368: 427–433.

Godovac-Zimmermann J et al. 1988a. Microanalysis of the aminoacid sequence of monomeric beta-lactoglobulin I from donkey(Equus asinus) milk. The primary structure and its homologywith a superfamily of hydrophobic molecule transporters. BiolChem Hoppe Seyler 369: 171–179.

Godovac-Zimmermann J et al. 1988b. The primary structure ofdonkey (Equus asinus) lysozyme contains the Ca(II) bindingsite of alpha-lactalbumin. Biol Chem Hoppe Seyler 369: 1109–1115.

Godovac-Zimmermann J et al. 1990. Covalent structure of the minormonomeric beta-lactoglobulin II component from donkey milk.Biol Chem Hoppe Seyler 371: 871–879.

Goldman AS et al. 1963. Milk allergy. 1. Oral challenge with milkand isolated milk proteins in allergic children. Pediatrics 32:425–443.

Gottschalk M et al. 2003. Protein self-association in solution:the bovine β-lactoglobulin dimmer and octamer. Protein Sci12:2404–2411.

Green MR, Pastewka JV. 1978. Lactoferrin is a marker for pro-lactin response in mouse mammary explants. Endocrinology 103:1510–1513.

Guo HY et al. 2007. Composition, physicochemical properties, ni-trogen fraction distribution, and amino acid profile of donkeymilk. J Dairy Sci 90: 1635–1643.

Halliday JA et al. 1991. The complete amino acid sequence offeline β-lactoglobulin II and a partial revision of the equineβ-lactoglobulin II sequence. Biochimi et Biophys Acta 1077:25–30.

Hambræus L. 1982. Nutritional aspects of milk proteins. In: PF Fox(ed.) Developments in Dairy Chemistry, vol 1: Proteins. AppliedScience Publishers, London, pp. 289–313.

Hambræus L. 1984. Human milk composition. Nutrit Abstr Rev.Series A 54: 219–236.

Haridas M et al. 1995. Structure of human diferric lactoferrin re-fined at 2.2 Å resolution. Acta Crystallographica D 51: 629–646.

Herrouin M et al. 2000. New genetic variants identified in donkey’smilk whey proteins. JProtein Chem 19: 105–115.

Herskovits TT. 1966. On the conformation of caseins. Optical rota-tory properties. Biochem 5: 1018–1026.

Hill DJ. 1994. Cow milk allergy within the spectrum of atopicdisorders. Clin Exp Allergy 24: 1137–1143.

Hill DJ, Hosking CS. 1996. Cow milk allergy in infancy and earlychildhood. J Allergy Clin Immunol 97: 9–15.

Holt C, Horne DS. 1996. The hairy casein micelle: evolution of theconcept and its implications for dairy technology. Nether MilkDairy J 50: 85–111.

Holt C, Jenness R. 1984. Interrelationships of constituents and par-tition of salts in milk samples from eight species. Comp BiochemPhysiol 77A: 175–282.

Holt C, Sawyer L. 1993. Caseins as rheomorphic proteins: inter-pretation of primary and secondary structures of the αs1-, β- andκ-caseins. J Chem Soc, Faraday Trans 89: 2683–2692.

Horne DS. 1998. Casein interactions: casting light on the blackboxes, the structure in dairy products. Intern Dairy J 8: 171–177.

Horne DS. 2003. Ethanol stability. In: PF Fox, PLH McSweeney(eds.) Advanced Dairy Chemistry, vol. 1: Proteins, 3rd edn.Kluwer Academic/Plenum Publishers, New York, pp. 975–1000.

Horne DS. 2006. Casein micelle structure: Models and muddles.Curr Opin Colloid Interface Sci11: 148–153.

Horrobin DF. 2000. Essential fatty acid metabolism and its modifi-cation in atopic eczema. Am J Clin Nutr71: 367S–372S.

Høst A. 1988. A prospective study of cow’s milk allergy in exclu-sively breast-fed infants. Incidence, pathogenetic role of earlyinadvertent exposure to cow’s milk formula, and characterizationof bovine milk protein in human milk. Acta Paediatrica Scandi-navica 77: 663–670.

Høst A. 1991. Importance of the first meal on the development ofcow’s milk allergy and intolerance. Allergy Proc 10: 227–232.

Humbert G et al. 2005. Plasmin activity and plasminogen activationin equine milk. Milchwissenschaft 60: 134–137.

Hurley WL. 2003. Immunoglobulins in mammary secretions. In:PF Fox, PLH McSweeney (eds.) Advanced Dairy Chemistry, vol1: Proteins, 3rd edn. Kluwer Academic/Plenum Publishers, NewYork, pp. 422–447.

Iacono G et al. 1992. Use of ass’ milk in multiple food allergy.J Pediatr Gastroenterol Nutr 14: 177–181.

Iametti S et al. 2001. Primary structure of κ-casein isolated frommares’ milk. J Dairy Res 68: 53–61.

International Dairy Foundation (IDF). 2009. The World Dairy Sit-uation 2009. Bulletin of the International Dairy Federation No

438. International Dairy Federation, Brussels, Belgium.Innis S. 2007. Fatty acids and early human development. Early Hum

Dev 83: 761–766.Izraely H et al. 1989. Factors determining the digestive efficiency of

the domesticated donkey (Equus asinus asinus). Q J Exp Physiol74: 1–6.

Jahreis G et al. 1999. The potential anticarcinogenic conjugatedlinoleic acid, cis-9, trans-11 C18:2, in milk of different species:cow, goat, ewe, cow, mare, women. Nutr Res 19: 1541–1549.

Jakobsson I, Lindberg T. 1978. Cow’s milk as a cause of infantilecolic in breast-fed infants. Lancet 11: 437–446.

Jakobsson I et al. 1985. Dietary bovine beta lactoglobulin is trans-ferred to human milk. Acta Paediatrica Scandinavia 74: 342–345.

P1: SFK/UKS P2: SFK


526 Part 4: Milk

Jarvinen KM et al. 2001. IgE and IgG binding epitopes on α-lactalbumin and β-lactoglobulin cow’s milk allergy. Int ArchAllergy Immunol 126: 111–118.

Jasinska B, Jaworska G. 1991. Comparison of structures of micellarcaseins of milk of cows, goats and mares with human milk casein.Anim Sci Pap Rep 7: 45–55.

Jauregui-Adell J. 1975. Heat stability and reactivation of mare milklysozyme. J Dairy Sci 58: 835–838.

Jenkins JA et al. 2007. Evolutionary distance from human homologsreflects allergenicity of animal food proteins. J Allergy Clin Im-munol 120: 1399–1405.

Jenness R. 1986. Lactational performance of various mammalianspecies. J Dairy Sci 69: 869–885.

Johnsen LB et al. 1995. Characterization of three types of humanαs1-casein mRNA transcripts. Biochem J 309: 237–242.

Kalliala H et al. 1951. On the use of mare’s milk in infant feeding.Acta Paediatrica 40: 97–117.

Kanamori M et al. 1981. Presence of O-glycosidic linkage throughserine residue in kappa-casein component from bovine maturemilk. J Nutr Sci Vitaminol (Tokyo) 27: 231–241.

Kanbe M. 1992. Traditional fermented milk of the world. In:Y Nakazawa, A Hosono (eds.) Functions of Fermented Milk:Challenges for the Health Sciences. Elsevier Applied Science,London.

Kaneko K et al. 2004. Bovine subclinical mastitis diagnosed on thebasis of amyloid A protein in milk. J Jpn Vet Med Assoc 57:515–518.

Keenan TW, Mather IH. 2006. Intracellular origin of milk fat glob-ules and the nature of the milk fat globule membrane. In: PFFox, PLH McSweeney (eds.) Advanced Dairy Chemistry, vol 2:Lipids, 3rd edn. Springer, New York, pp. 137–171.

Kilshaw PJ, Cant AJ. 1984. The passage of maternal dietary pro-teins into human breast milk. Int Arch Allergy Appl Immunol 75:8–15.

Koshiba T et al. 2000. Structure and thermodynamics of the ex-traordinary stable molten globule state of canine milk lysozyme.Biochem 39: 3248–3257.

Koshiba T et al. 2001. Energetics of three-state unfolding of a pro-tein: canine milk lysozyme. Protein Eng 14: 967–974.

Kotts C, Jenness R. 1976. Isolation of κ-casein-like proteins frommilk of various species. J Dairy Sci 59: 816–822.

Koroleva NS. 1991. Products prepared with lactic acid bacteriaand yeasts. In: RK Robinson (ed.) Therapeutic Properties ofFermented Milk. Elsevier Applied Science, Barking, pp. 159–179.

Kucukcetin A et al. 2003. Adaptation of bovine milk towards mares’milk composition by means of membrane technology for koumissmanufacture. Intern Dairy J 13: 945–951.

Kulisa M. 1977. The composition of mare’s milk in three horsebreeds with reference to N-acetylneuraminic acid. Acta Agrariaet Silvestria Series Zootechnica 27: 25–37.

Kunz C, Lonnerdal B. 1990. Casein and casein subunits in pretermmilk, colostrum, and mature human milk. J Pediatr GastroenterolNutr. 10: 454–461.

Kuroki R et al. 1989. Design and creation of a Ca2+ binding site inhuman lysozyme to enhance structural stability. Proc Natl AcadSci U S A 86: 6903–6907.

Kurmann JA et al. 1992. Encyclopedia of Fermented Fresh MilkProducts: An International Inventory of Fermented Milk, Cream,

Buttermilk, Whey, and Related Products. Van Nostrand Reinhold,New York.

Kyte J, Doolittle RF. 1982. A simple method for displayingthe hydropathic character of a protein. J Mol Biol 157: 105–132.

Langer P. 2009. Differences in the composition of colostrum andmilk in eutherians reflect differences in immunoglobulin transfer.J Mammal 90: 332–339.

Lara-Villoslada F et al. 2005. The balance between caseins andwhey proteins in cow’s milk determines its allergenicity. J DairySci 88: 1654–1660.

Larson BL. 1979. Biosynthesis and secretion of milk proteins: areview. J Dairy Res 46: 161–174.

Larson MA et al. 2003. Human serum amyloid A3 peptide enhancesintestinal MUC3 expression and inhibits EPEC adherence. Bio-chemi et Biophys Res Comm 300: 531–540.

Lauer BH, Baker BE. 1977. Amino acid composition of caseinisolated from the milks of different species. Can J Zool 55: 231–236.

LeBlanc MM et al. 1986. Relationships among serum immunoglob-ulin concentration in foals, colostral specific gravity, and colostralimmunoglobulin concentration. J Am Vet Med Assoc 189: 57–60

Leitch EC, Willcox MD. 1999. Elucidation of the antistaphylococcalaction of lactoferrin and lysozyme. J Med Microbiol 48: 867–871.

Lenasi T et al. 2003. Characterization of equine cDNA sequencesfor αs1-, β- and κ-caseins. J Dairy Res 70: 29–36.

Levine MA. 1998. Eating horses: the evolutionary significance ofhippophagy. Antiquity 72: 90–100.

Levy J. 1998. Immunonutrition: the pediatric experience. Nutrition14: 641–647.

Lonnerdal B. 2003. Lactoferrin. In: PF Fox, PLH McSweeney(eds.) Advanced Dairy Chemistry, vol 1: Proteins, 3rd edn.Kluwer Academic/Plenum Publishers, New York, pp. 449–460.

Lopez-Fandino R et al. 2006. Physiological, chemical and tech-nological aspects of milk-protein-derived peptides with anti-hypertensive and ACE-inhibitory activity. Intern Dairy J 16:1277–1293.

Lozovich S. 1995. Medical uses of whole and fermented mare milkin Russia. Cultured Dairy Products 30: 18–21.

Lucey JA, Singh H. 2003. Acid coagulation of milk. In: PF Fox,PLH McSweeney (eds.) Advanced Dairy Chemistry, vol 1: Pro-teins, 3rd edn. Kluwer Academic/Plenum Publishers, New York,pp. 1001–1026.

Lyster RLJ. 1992. Effect of calcium on the stability of mares’ milklysozyme. J Dairy Res 59: 331–338.

MacDonald D. 2001. The Encyclopedia of Mammals. Barnes andNoble/Andromeda Oxford Ltd., Abingdon.

Mack DR et al. 2003. The conserved TFLK motif of mammary-associated serum amyloid A3 is responsible for up-regulationof intestinal MUC3 mucin expression in vitro. Pediatr Res 53:137–142.

Madureira AR et al. 2007. Bovine whey proteins – Overviewon their main biological properties. Food Res Intel 40: 1197–1211.

Malacarne M et al. 2000. Observations on percentage distribu-tion of the main mare milk caseins separated by reversed-phase

P1: SFK/UKS P2: SFK



HPLC. Annali della Facolta di Medicina Veterinaria, Universitadi Parma (Italy) 20: 143–152.

Malacarne M et al. 2002. Protein and fat composition of mare’s milk:some nutritional remarks with reference to human and cow’s milk.Intern Dairy J 12: 869–897.

Malkoski M et al. 2001. Kappacin, a novel antibacterial peptide frombovine milk. Antimicrob Agents Chemother 45: 2309–2315.

Marchand S et al. 2009. Thermal inactivation kinetics of alkalinephosphatase in equine milk. Intern Dairy J 19: 763–767.

Marconi E, Panfili G. 1998. Chemical composition and nutritionalproperties of commercial products of mare milk powder. J FoodCompost Anal 11: 178–187.

Mariani P et al. 2001. Physicochemical properties, gross compo-sition, energy value and nitrogen fractions of Halflinger nursingmare milk throughout 6 lactation months. Anim Res 50: 415–425.

Martin P et al. 1996. The gene encoding αs1-casein is expressedin human mammary epithelial cells during lactation. Lait 76:523–535.

Martin P et al. 2011. Inter-species comparison of milk proteins:quantitative variability and molecular diversity. In: J Fuquay et al.(eds.) Encyclopedia of Dairy Sciences, 2nd edn. Elsevier, Oxford,3: 821–842.

Martuzzi F, Doreau M. 2006. Mare milk composition: recent find-ings about protein fractions and mineral content. In: N Margilia,W Martin-Rosset (eds.) Nutrition and Feeding of the Broodmare.EAAP publication No.20, Wageningen Academic Publishers,The Netherlands, pp. 65–76.

Masson PL, Heremans JF. 1971. Lactoferrin in milk from differentspecies. Comp Biochem Physiol 39B: 119–129.

Mateos A et al. 2009a. Equine αs1-casein: Characterization of al-ternative splicing isoforms and determination of phosphorylationlevels. J Dairy Sci 92: 3604–3615.

Mateos A et al. 2009b. Two-dimensional cartography of equine β-casein variants achieved by isolation of phosphorylation isoformsand control of the deamidation phenomenon. J Dairy Sci 92:2389–2399.

Mather IH. 2000. A review and proposed nomenclature for ma-jor proteins of the milk fat globule membrane. J Dairy Sci 83:203–247.

McDonald TL et al. 2001. Elevated extra hepatic expression andsecretion of mammary-associated serum amyloid A 3 (M-SAA3)into colostrum. Vet Immunol Immunopathol 3: 203–211.

McIntosh GH. 1996. Probiotics and colon cancer prevention. AsiaPac J Clin Nutr 5: 48–52.

McSweeney PLH, Fox PF. 2009. Advanced Dairy Chemistry, vol 3:Lactose, Water, Salts and Minor Constituents, 3rd edn. SpringerScience, New York.

Mercier J-C et al. 1976. Comparative study of the amino acid se-quences of the caseinomacropeptides from seven species. FEBSLetters 72: 208–214.

Mercier JC. 1986. Genetic engineering applied to milk produc-ing animals: some expectations. In: C Smith (ed.) ExploitingNew Technologies in Animal Breeding: Genetic DevelopmentsProceedings of a Seminar in the CEC Animal Husbandry Re-search Programme, Oxford University Press, Edinburgh, pp. 122–131.

Michaelidou A, Steijns J. 2006. Nutritional and technological as-pects of minor bioactive components in milk and whey: Growth

factors, vitamins and nucleotides. Intern Dairy J 16: 1421–1426.

Miclo L et al. 2007. The primary structure of a low-Mr multi-phosphorylated variant of β-casein in equine milk. Proteomics 7:1327–1335.

Minkiewicz P et al. 1993. The contribution of N-acetylneuraminicacid on the stabilization of micellar casein. Pol J Food Nutr Sci2: 39–48.

Miranda G et al. 2004. Proteomic tools to characterize the proteinfraction of Equidae milk. Proteomics 4: 2496–2509.

Molenaar AJ et al. 1996. Elevation of lactoferrin gene expressionin developing, ductal, resting, and regressing parenchymal ep-ithelium of the ruminant mammary gland. J Dairy Sci 79: 1198–1208.

Montagne P et al. 1998. Microparticle-enhanced nephelometric im-munoassay of lysozyme in milk and other human body fluids.Clin Chem 44: 1610–1615.

Monti G et al. 2007. Efficacy of donkey’s milk in treating highlyproblematic cow’s milk allergic children : an in vivo and in vitrostudy. Pediatr Allergy Immunol 18: 258–264.

Moore SA et al. 1997. Three dimensional structure of diferric bovinelactoferrin at 2.8 Å resolution. J Mol Biol 274: 222–236.

Morozova-Roche L. 2007. Equine lysozyme: the molecular basis offolding, self-assembly and innate amyloid toxicity. FEBS Letters581: 2587–2592.

Nakamura T et al. 2001. Occurrence of an unusual phosphorylatedN-acetyllactosamine in horse colostrum. Biochimi et Biophy Acta1525: 13–18.

Nakazawa T, Hosono A. 1992. Functions of Fermented Milk:Challenges for the Health Sciences. Elsevier Applied Science,London.

Nakhasi HL et al. 1984. Expression of κ-casein in normal andneoplastic rat mammary gland is under the control of prolactin.J Biol Chem 259: 14894–14898.

Natale M et al. 2004. Cow’s milk allergens identification by two-dimensional immunoblotting and mass spectrometry. Mol NutrFood Res 48: 363–369.

Neseni R et al. 1958. Milchleistung und Milchzusammenstzungvon Studen im Verlauf der Laktation. Archiv fur Tierzucht 1:191–129.

Neville MC. 2009. Introduction: alpha-lactalbumin, a multifunc-tional protein that specifies lactose synthesis in the Golgi. J Mam-mary Gland Biol Neoplasia 14: 211–212.

Nitta K et al. 1987. The calcium-binding properties of equinelysozyme. FEBS Letters 223: 405–408.

Nonaka Y et al. 2008. Spontaneous asparaginyl deamidation ofcanine milk lysozyme under mild conditions. Proteins 72: 313–322.

O’Connell JE, Fox PF. 2003. Heat-induced coagulation of milk. In:PF Fox, PLH McSweeney (eds.) Advanced Dairy Chemistry, vol.1: Proteins, 3rd edn. Kluwer Academic/Plenum Publishers, NewYork, pp. 879–945.

Ochirkhuyag B et al. 2000. Characterization of mare caseins. Iden-tification of αs1- and αs2-caseins. Lait 80: 223–235.

Oftedal OT et al. 1983. Lactation in the horse: milk compositionand intake by foals. J Nutr 113: 2096–2106.

Oftedal OT, Iverson SJ. 1995. Comparative analysis of non-humanmilks. In: RG Jensen (ed.) Handbook of milk composition. Aca-demic Press, California, pp. 749–827.

P1: SFK/UKS P2: SFK


528 Part 4: Milk

Ono T et al. 1989. Subunit components of casein micelles frombovine, ovine, caprine and equine milk. J Dairy Res 56: 61–68.

Outram AK. 2009. The earliest horse harnessing and milking. Sci-ence 323: 1332–1335.

Pagliarini E et al. 1993. Chemical and physical characteristics ofmare’s milk. Ital J Food Sci 4: 323–332.

Palmquist DL. 2006. Milk fat: origin of fatty acids and influence ofnutritional factors thereon. In: PF Fox, PLH McSweeney (eds.)Advanced Dairy Chemistry, vol 2: Lipids, 3rd edn. Springer, NewYork, pp. 43–92.

Park YW et al. 2006. Mare milk. In: YW Park, FW Haenlein (eds.)Handbook of Non-bovine Mammals. Blackwell Publishing, Ox-ford, pp. 275–296.

Parodi PW. 1982. Positional distribution of fatty acids in triglyc-erides from milk of several species of mammals. Lipids 17:437–442.

Patton S. 1999. Some practical implications of the milk mucins. JDairy Sci 82: 1115–1117.

Paulson M, Dejmek P. 1990. Thermal denaturation of whey pro-teins in mixtures with caseins studied by differential scanningcalorimetry. J Dairy Sci 73: 590–600.

Pelizzola V et al. 2006. Chemical-physical characteristics and fattyacid composition of mare’s milk. Milchwissenschaft 61: 33–36.

Perez MD, Calvo M. 1995. Interaction of β-lactoglobulin withretinol and fatty acids and its role as a possible biological functionfor this protein: A review. J Dairy Sci 78: 978–988.

Perez MD et al. 1993. Comparison of the ability to bind ligandsof β-lactoglobulin and serum albumin from ruminant and non-ruminant species. J Dairy Res 60: 55–63.

Phadungath C. 2005. Casein micelle structure: a concise review.Songklanakarin J Sci Technol 27: 201–212.

Phelan M et al. 2009. Casein-derived bioactive peptides: Biologicaleffects, industrial uses, safety aspects and regulatory status. InternDairy J 19: 643–654.

Piccione G et al. 2008. Daily rhythmicity in nutrient content ofasinine milk. Livestock Sci 116: 323–327.

Pieszka M, Kulisa M. 2005. Magnesium content in mares’ milkand growth parameters of their foals. J Elementology 10: 985–990.

Pikul J, Wojtowski J. 2008. Fat and cholesterol content and fatty acidcomposition of mares’ colostrums and milk during five lactationmonths. Livestock Sci 113: 285–290.

Pisano A et al. 1994. Characterization of O-linked glycosylationmotifs in the glycopeptide domain of bovine κ-casein. Glycobi-ology 4: 837–844.

Plowman JE et al. 1999. Structural features of a peptide correspond-ing to human κ-casein residues 84–101 by 1H-nuclear magneticresonance spectroscopy. J Dairy Res 66: 53–63.

Qi PX. 2007. Studies of casein micelle structure: the past and thepresent. Lait 87: 363–383.

Rainhard P. 1986. Bacteriostatic activity of bovine milk lactoferrinagainst mastitic bacteria. Veter Microbiol 11: 387–392.

Ramos JJ et al. 1994. Vitamin B12 levels in ewe colostrum and milkand in lamb serum. Vet Res 25: 405–409.

Rasmussen LK et al. 1995. Human αs1-casein: purification andcharacterization. Comp Biochem PhysiolB 111: 75–81.

Rassin DK et al. 1978. Taurine and other free amino acids in milkof man and other mammals. Early Hum Dev 2/1: 1–13.

Restani P et al. 1995. Evaluation by SDS-PAGE and immunoblottingof residual antigenicity in hydrolysed protein formulas. Clin ExpAllergy 25: 651–658.

Recio I et al. 2009. Bioactive components in sheep milk. In: Y.W.Park (ed.), Bioactive Components in Milk and Dairy Products.Wiley-Blackwell Publishers. Ames, IA, pp. 83–104.

Restani P et al. 2002. Cross-reactivity between mammalian proteins.Ann Allergy Asthma Immunol 89: 11–15.

Rudloff S, Kunz C. 1997. Protein and nonprotein nitrogen compo-nents in human milk, bovine milk, and infant formula: Quanti-tative and qualitative aspects in infant nutrition. J Pediatr Gas-troenterol Nutr 26: 328–344.

Saito T, Itoh T. 1992. Variations and distributions ofO-glycosidically linked sugar chains in bovine κ-casein. J DairySci 75: 1768–1774.

Salimei E. 2011. Animals that produce dairy foods: Donkey. In: W.Fuquay et al. (eds.) Encyclopedia of Dairy Sciences, vol 1, 2ndedn. Elsevier. Oxford, pp. 365–373.

Salimei E et al. 2002. Major constituents leptin, and non-proteinnitrogen compounds in mares’ colostum and milk. Reprod NutrDev 42: 65–72.

Salimei E et al. 2004. Composition and characteristics of ass’ milk.Anim Res 53: 67–78.

Sarwar G et al. 1998. Free amino acids in milk of human subjects,other primates and non-primates. Brit J Nutr 79: 129–131.

Savage RJG, Long MR. 1986. Mammalian Evolution: an IllustratedGuide. Natural History Museum, London.

Savalle B et al. 1988. Modelization of gastric digestion of milk pro-teins. In: CA Barth, E Schlimme (eds.) Milk Proteins. Nutritional,Clinical, Functional and Technological Aspects. Steinkopff Pub-lishers, Darmstad, pp. 143–149.

Savilahti E, Kuitunen M. 1992. Allergenicity of cow milk proteins.J Pediatr 121: S12–S20.

Sawyer L. 2003. β-Lactoglobulin. In: PF Fox, PLH McSweeney(eds.) Advanced Dairy Chemistry, vol. 1: Proteins, 3rd edn.Kluwer Academic/Plenum Publishers, New York, pp. 319–386.

Schaafsma G. 2003. Nutritional significance of lactose and lactosederivatives. In: H Roginski, JW Fuquay, PF Fox (eds.) Ency-clopedia of Dairy Science. Academic Press, London, pp. 1529–1533.

Schmidt DG et al. 1995. Raising the pH of the pepsin-catalyzedhydrolysis of bovine whey proteins increases the antigenicity ofthe hydrolysates. Clin Exp Allergy 25: 1007–1017.

Schryver HF et al. 1986. Lactation in the horse: the mineral com-position of mare milk. J Nutr 116: 2142–2147.

Sellnow L. 2006. Food Factory. Article 10, Equine Anatomy andPhysiology. Available at http://www.thehorse.com/pdf/anatomy/anatomy10.pdf. Accessed on October 3, 2010.

Selo I et al. 1999. Allergy to bovine β-lactoglobulin: specificityof human IgE to tryptic peptides. Clin Exp Allergy 29: 1055–1063.

Shah NP. 2000. Effects of milk-derived bioactives: an overview.Brit J Nutr 84: S3–S10.

Shekar CP et al. 2006. κ-Casein-deficient mice fail to lactate. ProcNatl Acad Sci U S A 103: 8000–8005.

Shimazaki KI et al. 1994. Comparative study of the iron-bindingstrengths of equine, bovine and human lactoferrins. J Dairy Res61: 563–566.

P1: SFK/UKS P2: SFK



Silva SV, Malcata FX. 2005. Caseins as source of bioactive peptides.Intern Dairy J 15: 1–15.

Simopoulos AP. 2002. The importance of the ratio of omega-6/omega-3 essential fatty acids. Biomedical Pharmacotherapy56: 365–379.

Singh H et al. 1997. Physico-chemical properties of milk. In: PFFox (ed.) Advanced Dairy Chemistry vol. 3: Lactose, Water, Saltsand Vitamins, 2nd edn. Chapman and Hall, London, pp. 469–518.

Smith MC, Sherman DM. 2009. Goat Medicine, 2nd edn. Wiley-Blackwell, New York.

Smolders EAA et al. 1990. Composition of horse milk during thesuckling period. Livest Prod Sci 25: 163–171.

Sneddon JC, Argenzio RA. 1998. Feeding strategy and water home-ostasis in equids: the role of the hind-gut. J Arid Envir 38:493–509.

Sneddon JC et al. 2006. Mucosal surface area and fermentationactivity in the hind gut of hydrated and chronically dehydratedworking donkeys. J Anim Sci 84: 119–124.

Solaroli G et al. 1993. Composition and nutritional quality of mare’smilk. Ital J Food Sci 5: 3–10.

Sood SM, Slattery CW. 2000. Association of the quadruply phos-phorylated β-casein from human milk with the nonphosphory-lated form. J Dairy Sci 83: 2766–2770.

Sørensen ES et al. 2003. The phosphorylation pattern of humanαs1-casein is markedly different from the ruminant species. EurJ Biochem 270: 3651–3655.

Souci SW et al. 2000. Food Composition and Nutrition Tables, 6thedn. Medpharm Scientific Publishers, Stuttgart.

Spencer A et al. 1999. Expression, purification, and characterizationof the recombinant calcium-binding equine lysozyme secreted bythe filamentous fungus Aspergillus niger: comparisons with theproduction of hen and human lysozymes. Protein Expression andPurification 16: 171–180.

Spitzauer S. 1999. Allergy to mammalian proteins: at the border-line between foreign and self? Int Arch Allergy Immunol 120:259–269.

Stasiuk SX et al. 2000. Cloning of a marsupial kappa-casein cDNAfrom the brushtail possum (Trichosurus vulpecula). Reprod FertilDev 12: 215–222.

Stoyanova LG et al. 1988. Freeze-dried mares’ milk and its poten-tial use in infant and dietetic food products. Voprosy Pitaniia 2:64–67.

Stromqvist M et al. 1995. Human milk β-casein and inhibition ofHelicobacter pylori adhesion to human gastric mucosa. J PediatrGastroenterol Nutr. 21: 288–296.

Stuart CA et al. 1984. Passage of cows’ milk protein in breast milk.Clin Allergy 14: 533–535.

Surono IS, Hosono A. 2002. Fermented milks: Types and standardsof identity. In: H Roginski, JA Fuquay, PF Fox (eds.) Ency-clopedia of Dairy Sciences. Academic Press, London, UK, pp.1018–1069.

Swaisgood HF. 2003. Chemistry of caseins. In: PF Fox, PLHMcSweeney (eds.) Advanced Dairy Chemistry, vol 1: Pro-teins, 3rd edn. Kluwer Academic/Plenum Publishers, New York,pp. 139–202.

Tada M et al. (2002). Stabilization of protein by replacementof a fluctuating loop: Structural analysis of a chimera ofbovine α-lactalbumin and equine lysozyme. Biochem 41: 13807–13813.

Tamine AY, Marshall VME. 1984. Microbiology and technologyof fermented milk. In: FL Davies, BA Law (eds.) Advances inthe Microbiology and Biochemistry of Cheese and FermentedMilk. Elsevier Applied Science Publishers, London, pp. 118–122.

Tamine AY, Robinson RK. 1999. Yoghurt: Science and Technology.Woodhead Publishing Ltd., Cambridge.

Taylor SL. 1986. Immunologic and allergic properties of cows’ milkproteins in humans. J Food Prot 49: 239–250.

Thoma-Worringer C et al. 2006. Health effects and technologi-cal features of caseinomacropeptide. Intern Dairy J 16: 1324–1333.

Trujillo JM et al. 1962. Chromosomes of the horse, the donkey, andthe mule. Chromosoma 13: 243–248.

Tsioulpas A et al. 2007. Changes in physical properties of bovinemilk from the colostrum period to early lactation. J Dairy Sci 90:5012–5017.

Tsuge H et al. 1992. Crystallographic studies on a calcium-bindinglysozyme from equine milk at 2.5 Å resolution. J Biochem 111:141–143.

Turner AW. 1945. Digestibility of milk as affected by various typesof treatment. Food Res Inter 10: 52–59.

Ullrey DE et al. 1966. Composition of mare’s milk. J Anim Sci 25:217–222.

Ullrey DE et al. 1974. Iron, zinc and copper in mare’s milk. J AnimSci 38: 1276–1277.

Uniacke-Lowe T. 2011. Studies on equine milk and comparativestudies on equine and bovine milk systems. PhD Thesis. TheNational University of Ireland, Cork.

Uniacke T, Fox PF. 2011. Equid milk. In: J Fuquay et al. (eds.).Encyclopedia of Dairy Sciences, vol 3, 2nd edn. Elsevier, Oxford,pp. 518–529.

Uniacke T et al. 2010. Equine milk proteins: Chemistry, structureand nutritional significance. Intern Dairy J 20: 609–629.

Urashima T et al. 1989. Structural determination of three neutraloligosaccharides obtained from horse colostrum. Carbohydr Res194: 280–287.

Urashima T et al. 2001. Oligosaccharides of milk and colostrum innon-human mammals. Glycoconj J 18: 357–371.

Urashima T et al. 2009. Milk oligosaccharides. In: PLH Mc-Sweeney, PF Fox (eds.) Advanced Dairy Chemistry, vol. 3: Lac-tose, Water, Salts and Minor Constituents, 3rd edn. Springer,London, pp. 295–349.

Van Hooydonk ACM et al. 1984. Kinetics of the chymosin-catalysedproteolysis of κ-casein in milk. Neth Milk Dairy J 38: 207–222.

Vesa TH et al. 2000. Lactose intolerance. J Am Coll Nutr 19:165S–175S.

Vila C et al. 2001. Widespread origins of domestic horse lineages.Science 291: 474–477.

Vincenzetti S et al. 2008. Donkey’s milk protein fractions charac-terization. Food Chem 106: 640–649.

Visser S et al. 1982. Isolation and characterization of β- andγ -caseins in horse milk. Biochem J 203: 131–139.

Vreeman HJ et al. 1986. Characterization of bovine κ-casein frac-tions and the kinetics of chymosin-induced macropeptide releasefrom carbohydrate-free and carbohydrate-containing fractionsdetermined by high-performance gel-permeation chromatogra-phy. Biochem J 240: 87–97.

P1: SFK/UKS P2: SFK


530 Part 4: Milk

Waelchli RO et al. 1990. Relationships of total protein, specificgravity, viscosity, refractive index and latex agglutination to im-munoglobulin G concentration in mare colostrum. Equine Vet J22: 39–42.

Wal J-M. 2002. Cow’s milk proteins/allergens. Ann Allergy AsthmaImmunol 89: 3–10.

Wal J-M. 2004. Bovine milk allergenicity. Ann Allergy AsthmaImmunol 93: S2–S11.

Walstra P. 1990. On the stability of casein micelles. J Dairy Sci 73:1965–1979.

Walstra P, Jenness R. 1984. Dairy Chemistry and Physics. JohnWiley & Sons, New York.

Walstra P et al. 2006a. Dairy Science and Technology. CRC Press,Boca Raton, FL.

Walstra P et al. 2006b. Dairy Technology:Principles of Milk Prop-erties and Processes. CRC/Taylor Francis, New York.

Welsch U et al. 1988. Structural, histochemical and biochemicalobservations on horse milk-fat-globule membranes and caseinmicelles. Histochemistry 88: 357–365.

Whigham LD et al. 2000. Conjugated linoleic acid: implications forhuman health. Pharmacol Res 42: 503–510.

Widdowson EM. 1984. Lactation and feeding patterns in differentspecies. In: DLM Freed (ed.) Health Hazards of Milk. BailliereTindall, London, pp. 85–90.

Winter P et al. 2006. The potential of measuring serum amyloidA in individual ewe milk and in farm bulk milk for monitor-ing udder health on sheep dairy farms. Res Vet Sci 81: 321–326.

Wong NP et al. (eds.) 1988. Fundamentals of Dairy Chemistry, 3rdedn. Van Nostrand Reinhold, New York.

Yongfa W et al. 1984. Prevalence of primary adult lactose malab-sorption in three populations of northern China. Human Genetics67: 103–106.

Yoshikawa M et al. 1982. Chemical characteristics of bovine caseinmicelles fractionated by size on CPG-10/3000 chromatography.Agri Biol Chem 46: 1043–1048.

Zbikowska A et al. 1992. The influence of casein micelle size onselected functional properties of bulk milk proteins. Polish J FoodNutr Sci 1: 23–32.

Zicker SC, Lonnerdal B. 1994. Protein and nitrogen compositionof equine (Equus caballus) milk during early lactation. CompBiochem Physiol 108A: 411–421.

Documents

Food Biochemistry and Food Processing (Simpson/Food Biochemistry and Food Processing) || Equid Milk: Chemistry, Biochemistry and Processing